of sum and inquam. The sign of the first person singular is also wanting in the perfect indicative, and in the futures in -bo and -ro. The second person singular is represented by -8 in every case but one-that of the perfect indicative, which substitutes -sti. The third singular is always -t; the first plural always -mus; the second plural always -tis, except in the perfect indicative, when it is -stis, to correspond with the singular of the same person; and the third plural is always -nt, though this is occasionally dropt in the third person plural of the perfect indicative. If we may judge from the -to, -tote of the imperative, these person-endings must have been originally ablative or causative inflexions of the pronouns. The original form of the imperative suffix in the singular number was -tod or -tud, which is unequivocally an ablative inflexion (above, Chap. VIII. § 8).

The person-endings of the passive verb present some difficulties to the inquiring philologist. In fact, only the third person, singular and plural, seems to have been preserved free from mutilation or suppression. The terminations of the passive should, according to the rules of sound philology, present themselves as inflexions or cases of the active person-endings. If, then, we compare the active amat, amant, amare, with the corresponding passive forms, amatur, amantur, amarier, we must conclude that r, connected with the active form by a short vowel, e or u, is the sign of the passive voice, and that this amounts to an inflexion of the active form analogous to the adverbs in -ter (leni-ter, gnavi-ter, &c.), -tus (cali-tus, &c.), or -tim (gradatim, &c.). In fact, the isolated particle igi-tur supplies a perfect analogy for the passive person-endings -tur and -ntur. This particle, as we have seen (above, pp. 289, 335), is an extension in -tur from the composite form i-gi (cf. e-go, er-ga, e-ho, e-ja), and it has the locative meaning "thereupon" in a Fragment of the XII. Tables (above, p. 204). We have also seen that the adverbs in -ter, -tim are used in a locative sense. And whether we conclude that tur is a locative like Tót, or identical with -tus-Oev, and therefore bearing a locative meaning only as the act of separation implies proximity at the moment of separation (above, p. 330), there can be no doubt that it does bear that locative sense, which is required by the person-endings of the passive voice. The identity of -tur with -ter (-tim) is farther shown by the form amari-er, which stands by the side of ama-tur. According to

this, the first persons amor and amamur are contractions of amoměr, amāmusěr, according to the Sanscrit analogy (comp. bharé with pépouai, &c. New Crat. §§ 352, 362). The second persons, amaris (amare) and amamini, are altogether different forms; they seem to be two verbals, or participial nouns, of the same kind respectively as the Latin and Greek active infinitive, amare amase (compare dic-sis-se, es-se, Gr. yéλaïs, vчoïs, &c.), and the passive participle τυπτόμενος. The verbal, which stands for the second person singular of the passive verb, was probably, in the first instance, a verbal noun in -sis; compare πρᾶξις, μίμησις, &c. That which represents the second person plural is the plural of a form which is of very frequent occurrence in the Latin language (New Crat. § 362). The older form ended in -minor, and is preserved in the imperative, which in old Latin had a corresponding second person singular in -mino : thus we have antestamino (Legg. xii. Tab. I. Fr. 1, above, Ch. VI. § 7), famino (Fest. p. 87), præfamino (Cat. R. R. 135, 140), fruimino (Inscr. Grut.), for antestare, fare, præfare, fruere ; as well as arbitraminor (Plaut. Epid. V. 2, 30) and progrediminor (id. Pseud. III. 2, 70) for arbitramini and progredimini. The use of these verbals, with a fixed gender, and without any copula, to express passive predications referring to the second person, is one of the most singular features in the Latin language, and the former can only be compared to the Greek use of the infinitive to express the second person imperative.

§ 3. Doctrine of the Latin Tenses.

There is, perhaps, no one department of classical philology, in which so little has been done as in the analysis and simplification of the Latin tenses. They are still arranged and designated as they were in the beginning; and no one seems to have discerned the glaring errors inseparable from such a system. Even among the more enlightened, it is not yet agreed whether certain tenses are to be referred to the indicative or to the subjunctive mood, and forms of entirely different origin are placed together in the same category.

Without anticipating the discussion of the difficulties which beset the doctrine of the Latin tenses, I will premise that, practically, the regular verb has four moods and five tenses, which are known by the following names, and represented, in my

Grammar, by the notation attached to the terminology: the indicative (A), imperative (B), subjunctive (C), and infinitive (D) moods, and the present (I), imperfect (II), perfect (III), pluperfect (IV), and future (V) tenses. Thus, to avoid repeating the names, A. III. will represent the present indicative, C. II. the imperfect subjunctive, and so on.

An accurate examination of all the forms in the Latin language will convince us that there are only two ways in which a tense can be formed organically from the root of a Latin verb. One is, by the addition of s-; the other, by the addition of i-. We find the same process in the Greek verb; but there it is regular and systematic, supplying us throughout with a complete series of primary and secondary, or definite and indefinite tenses'. In Greek, we say that the addition of σ- to the root forms the aorist and future, that the same adjunct in a more guttural form makes the perfect, and that the insertion of indicates the conjunctive or optative mood. Moreover, we have in the Greek verb an augment, or syllable prefixed for the purpose of marking past time as such, and traces at least of the systematic employment of reduplication to designate the continuance of an action. As the ancient epic poetry of the Greeks neglects the augment, we may understand how it fell into desuetude among the Romans. The reduplication too, though common to all the old Italian languages, is of only partial application in the existing forms of the Latin verb. With regard to the value of the tenses in σand-, the same holds to a certain extent in Latin also; but while the principle is here susceptible of a double application, it is, on the other hand, interrupted by the operation of a system

1 For the convenience of the reader, I will repeat here the distinctions which I have elsewhere quoted from J. L. Burnouf's Méthode pour étudier la Langue Grecque, pp. 215, sqq.

The Imperfect expresses simultaneity with reference to je lisais 1

The Aorist.

The Pluperfect.

posteriority . anteriority

some other time

je lus 2

j'avais lus.

après que vous eutes fini d'écrire.

3 avant que vous eussiez écrit.

of composite tenses which is peculiar to the Latin language, and still more so by the irregular use of the affix -s to express derived or indefinite tenses.

Before I proceed to examine the tense-system of the Romans, as it appears in all the complications of an ordinary verb, it will be as well to analyse, in the first instance, the substantive verb which enters so largely into all temporal relations.

The Latin language has two verbs signifying "to be:" one contains the root es-, Sanser. as-, Greek co-, Lith. es-; the other, the root fu-, Sancr. bhû-, Gr. qv-, Lith. bu-. The inflexions of es- are as follows:

Or locative of a verbal in -sis, expressing the action of the verb1.

es-se.

PARTICIPLE. E.

Nom. 'sen[t]s (in ab-sens, præ-sens, &c.) originally esen[t]s

Throughout the Latin verb we may observe, as in the case of ero here, that the element i has vanished from the first person of the future; for ero does not really differ from esum, the present indicative. The explanation of this may be derived from the fact, that in English the first and the other persons of the future belong to different forms: where an Englishman says, "I shall" of himself, he addresses another with "you will;" and conversely, where he asserts of another that "he shall," he tells him, "I will." The third person plural erunt is only another way of writing erint; u, being substituted, as it so frequently is, for is, to which the qualifying i had been ultimately reduced. But besides the form of the future in i, we have in old Latin another expression of it in the inchoative form esco for es-sco (Legg. xi. Tab. apud Gell. XX. i. Tab. 1. fr. 3: Lucret. I. 613: Festus, s. v. escit, p. 77; superescit, p. 302; nec, p. 162; obescet, p. 188: and Müller, Suppl. Annot. p. 386).

The verb fu-, which appears as a supplementary form or auxiliary tense of the substantive verb, is really a distinct verb,

1 New Crat. § 410.