1850.] Greek and Latin Verbs. 659 cation, as scandi, verti, tuli, (for which an older tetuli is likewise found) or by a contracting together of the radical and reduplicative syllables, as feci for fafici, vidi for vividi. All others are compound and contain the perfect of one or other of the substantive verbs, a part that of Es, as in scrip-si, duc-si, a part that of FU, as in ama-vi, mon-ui, (compare for the latter pot-ui, where the fact of the composition cannot for a moment be doubted, and where fui has suffered the same excessive mutilation). There are sundry peculiarities about the perfects of both languages, some of which are not easy to explain and are still made subjects of controversy among the philologists. The Greek has adopted an unchangeable union-vowel throughout. The x, (лɛлαídεv-x-a) and the aspiration, (néñoμ-q-a) characteristic of its so-called first perfect, are probably inorganic intrusions, and not significant. The first person singular in each verb has lost the personal ending, as has the Sanscrit also, and the final i of the Latin, as distinguished from the o of the present and future, is noteworthy. The ti of the Latin second person singular corresponds to the Sanscrit ending tha, which also is used only in this tense, and which the Greek has preserved only in a few isolated instances, as noda, oïda. The s by which it, as well as the ending of the second person plural, is preceded, is likewise an inorganic intrusion, and to be compared with the s appearing so often before the of the Greek medial terminations, (7v778-6-Dov, etc.) The runt of the third person plural is explained as a further composition with sunt; this person showing in more than one instance a tendency toward a great and unnecessary fulness of form. These three tenses, the present, imperfect, and perfect, are the only ones for which the common language in its earliest period struck out separate and original forms, and of which, therefore, the relics can be traced out in all the members of the family since their dispersion. But one of the chief relations of time, the future, still remains unprovided with its appropriate expression, and we have next to inquire how, and how far in a like spirit, the two languages we are considering respectively supplied this deficiency. It would appear to us easy to have formed a future by a method analogous to that adopted for the simple preterite, by prefixing another adverbial augment, which should point the mind forward instead of back to the time of the action. A future so constructed actually exists among some of the Slavonic dialects, but, for whatever reason, the mother tongue originated none such. Perhaps at first the present was found capable of performing both offices satisfactorily; there are many phenomena in language which prove an easy transition between present and future signification, and the Gothic and Anglo-Saxon, to the end of their existence, contented themselves with nothing better. But, the deficiency once felt and its remedy determined upon, that recourse should have been had to optative and conditional forms cannot appear otherwise than most natural to us, who say I shall go, that is, I ought to go, and I will go, that is, I wish to go. Of such origin are both the Greek and Latin futures, and therefore closely related in spirit, although the special means used by either language are diverse. The Greek tense is fully explained by the Sanskrit, with which it is identical. It is constructed by compounding with the verbal root a tense which has been developed from the optative of the root Es of the substantive verb. We may conceive that tense to have been, in its oldest and fullest form, oío, and its relation to the still subsisting optative (a) to be simply this, that the former adopted, or retained, the full primary endings, which belong to an unaugmented tense, while the latter went through the process which reduced those endings to a conformity with the general optative model. Starting from this oío, it is not difficult to explain and connect the at present apparently discordant phenomena presented by the Greek futures. The old Doric dialect, here as elsewhere, gives us the most ancient and least corrupted forms, and in such future persons as βοαθη-σίω πραγ-σίομες (πραξίομες), which have lost only the initial ɛ, furnishes the strongest evidence to be gathered within Greek territory of the correctness of the explanation we have given of the origin of the tense. Excepting in these few relics, the which was originally the life and soul of the formation, (being the root I, to wish, desire, go; the universal characteristic of the optative,) has disappeared; yet leaving behind it traces: on the one hand, in the so-called Doric future middle, where, changed to ɛ, it is contracted with the following union-vowel; as φευξούμαι, for φενξέομαι, for φενξίομαι; on the other hand, in a few Homeric futures, where it is assimilated to the preceding sibilant; as, ἀρέσσω, ὀλέσσω, for ἀρεσίω, etc. In the liquid verbs, the conjunction of a liquid and sibilant being offensive to Greek ears, the initial ɛ was retained, and then, after the had been dropped, and the sibilant too had disappeared in obedience to the rule already noted, this ɛ likewise became contracted with the union-vowel, giving the so-called second future; as, pevc, perõvμer, for μevéoμev, for μeveGouer. Finally, in the usual conjugation, both ɛ and were lost and the sibilant alone remained as sign of the tense. Of kindred origin with this oío, probably, is the Latin ero, it too having lost its optative sign, and made the customary change of s into r. But of this the Latin did not make use in constructing its compound futures; it chose rather here, as in the imperfect, to avail itself of the correspond ε ι ε เ 1850.] Greek and Latin Verbs. 661 ing form from the other root, viz. bo, and by its aid composed a tense which bears the same relation to that of the Greek as its imperfect to the Greek aorist: (monebo: doxow:: monebam: ¿dóxŋoα). It was, however, only the verbs of the first and second conjugation which the Latin was compelled thus throughout to supply with a compound tense; in the third, and generally in the fourth, it converted the simple present optative itself into a future. For the legêm, legês, legêt, of the Latin, are incontestably the λéyou-t, éyois, héyoi (7) of the Greek, the original diphthong ai made up of the union-vowel and the optative sign, having been by the one language changed in one of its constituents, according to a general analogy of that language, and, in the other, contracted into the diphthongal vowel é, by a rule no less general, and which the Sanskrit, also, in its tudés, tudét, tudéma, etc., has obeyed. But in the second conjugation, it was impossible that any tense so constructed should be distinguished from the present, and there accordingly the compound future became necessary. Before the analogous difficulty which necessitated a similar proceeding in the first conjugation can be fully explained, we must take into account the present subjunctive. The Greek subjunctive is plainly generated by a lengthening of, a pausing or dwelling upon, the union-vowel, as if to denote the doubtful or conditional nature of the relation subsisting between the action expressed by the root, and the actor signified by the personal ending. Nor is the Latin of different origin; legâmus, legâtis, stand in the same relation to legimus, legitis, as λέγωμεν, λέγητε τo λέγομεν, λέγετε ; it being recollected that the union-vowel was in both languages originally a, and that the Latin knows nothing more of the customary Greek changes of long a into and o, than of those of short a into ε and o. Proper subjunctives by descent, then, are the subjunctives present of the last three Latin conjugations. But in the first it would have been impossible to distinguish such a tense from the present indicative. The optative was therefore, made to do duty here as a subjunctive, and the same necessity for a compound future arose as in the second conjugation. There are a few other optatives performing in Latin the office of subjunctives, which should be noticed here, as deviating somewhat from the general model. They are sim, edim, velim, and the like. These are analogous in formation to the optatives of the Greek conjugation in u. To that conjugation, whose chief characteristic is the lack of a union-vowel, the verbs from which they are derived originally belonged, and in some of their persons, as has been already remarked, they yet give evidence of their relationship to it; compare es-t es-tis, vul-tis; for sim, sis, etc., are still found in the earlier authors siem, sies, etc., forms identical with the Greek -(o)ínv, ¿-(o)íns. The VOL. VII. No. 28. 56 others had lost before the historical period of the language the e with which siem parted during that period. It ought not to surprise us to find the Latin dealing thus arbitrarily with, and confounding, optative and subjunctive. As being both conditionals they are originally of very near kin, and it is only the Greek language that, giving one the primary, the other the secondary, endings, has succeeded in fully separating them. The Gothic has nothing but an optative, nor does the Sanskrit exhibit more than traces of a subjunctive, and that only in the Veda dialect. The two imperatives correspond as closely as possible; compare Léɣɛ, lege, Lɛyézw, legito, kézɛrɛ, legite, heyóvrov, legunto. In the second person singular, both have lost the ending 7 or 1, as has the Sanskrit also in the great majority of its verbs. It appears, however, in the Greek verbs in u, either in its full form, as rivɛti, dídovi, or corrupted to the sibilant, as in eo and doo. The other and more frequent form of the third person plural 2ɛyérwoar, is explained, like the corresponding person of the Latin perfect, as a composition with a form of the substantive verb. The additional forms of the Latin second persons, legito, (for legitot,) and legitote, have become obsolete not only in Greek, but also in the classical Sanskrit. The Veda dialect still preserves them. The imperative, it will be noticed, has no mode-sign. Its sufficient characteristic was the tone of voice in which a command is wont to be uttered. These are the only temporal and modal forms concerning which it can be proved that they ever were possessed in common by both languages. They are, it will easily be seen, far from constituting the whole verbal apparatus, being only the indicative, subjunctive, optative, and imperative of the present, and the indicatives of a preterite and perfect, and even among these we have had to point out in the Latin verb various perversions, losses, and substitutions, more or less complete. These are probably all that, in the state of culture which prevailed in the parent nation at the time of separation of the two tribes, had been found necessary in order to the conveyance with requisite clearness of their thoughts and ideas. So much is characteristic of the race; the rest is the work of individual nations, undertaken when called for by their sense of the deficiencies which still existed in their store of verbal forms, and variously executed according to their command over the resources of formation at the time when they became sensible of such deficiencies. The Sanskrit, unsurpassed in formative power, failed greatly in its sense of syntactical distinctions, and was even unsuccessful in maintaining unimpaired the shades of meaning properly belonging to the forms of which it was already in possession. 1850.] Greek and Latin Verbs. 663 The Teutonic was content with only a present and preterite, indicative and subjunctive, until the time for the production even of compound forms was past, and recourse was then necessarily had to periphrases. With both Latin and Greek the case is quite otherwise; each language has given birth to a number of new compound moods and tenses, thus filling up its verb to answer to its own notion of completeness. As to tenses, the correspondence between the two is very close, nor does the Greek, saving in its aorist, maintain any advantage over the Latin. We have already seen how each provided itself with a future. The future. perfect, combining the two temporal ideas denoted by its name, would naturally be constructed out of the two tenses expressing those ideas, that is to say, by appending the future of the substantive verb to the root of the perfect. Accordingly we find in Greek, out of nɛлaidev and ἔσομαι, πεπαιδεύσομαι constructed; in Latin out of cecin and ero, cecinero. As plainly is the pluperfect a combination of perfect and imperfect, and in this tense also is the coincidence of the two languages quite as close, although not as evident, as in the one last mentioned. The composition of cecineram no one can fail at once to perceive; that of nɛñaidɛúzɛi is not so readily apparent. Its origin from inɛnαideux-coα, however, can be historically established. The Homeric form is -zea, the σ having been dropped according to the rule so often already noticed; this by regular contraction becomes -x, the form in use among the older Attics, and finally, by an irregular contraction (which finds its parallel in Baoiλéas ßasihɛis) it assumed the form of -x, which afterward came to prevail throughout the whole inflection of the tense. The final of the first person singular is an inorganic addition of that letter, such as is often found occurring in Greek. The two languages, so closely accordant in their methods of expressing temporal relations, are as widely discordant in their whole system of moods. The Greek, with a keener sense of the nicer shades of syntactic relations and differences than any other language has displayed, with a most flexible system of sounds, and, in the series belonging to its present, a complete model after which to work, went on to create after its analogy that exuberant store of modal forms which distinguishes the Greek verb; the Latin, less critical, with less pliant materials to work upon, with more stubborn phonetic laws, and left destitute of a model, devised to fill out its subjunctive compositions with the simple subjunctive tenses of the substantive verb. Between the verbal nouns, the infinitives, attached to each verb respectively, no connection is to be pointed out. The Latin, in its supines, has preserved the formation of which the Sanskrit still makes |