A Short Historical Latin Grammar: Chapter VI: Verb

English Grammar Advanced 19th century W. M. Lindsay

Pages 86–117 of the book.

The I.-Eur. Verb had two Conjugations, (1) the Thematic, in which the Person-suffixes were attached to the Verb-root by means of a connecting Vowel, ĕ or ŏ (ch. x. § 11), called the Thematic Vowel, e.g. Gk. λέγ-ο-μεν, λέγ-ε-τε, where the suffix of the First Person Plural, -μεν, and the suffix of the Second Person Plural, -τε, are attached to the root of the Verb, λεγ-, by means of the Vowels ο and ε; (2) the Athematic, in which the Person-suffixes were attached directly to the Verb-root without the help of this connecting vowel, e.g. Gk. ἴ-μεν, ἴ-τε. The Thematic Conjugation had its 1 Sg. Pres. Ind. in -ō, and used the diphthong oi in its Optative, e.g. Gk. λέγ-ω, λέγ-οι-τε. The 1 Sg. Pres. Ind. of the Athematic Conjugation had the suffix -mī, and in the Optative īē was used in the Sing. Act., elsewhere ī, e.g. Gk. εἴην, ἱστα-ίη-ν, ἱστα-ῖ-μεν (ἱσταῖμεν). The so-called ‘irregularity’ of Latin Verbs like sum, eo, volo, is mainly due to the fact that they belonged to the I.-Eur. Athematic Conjugation. The root of the Verb sum is es- (cf. Inf. es-se), and es-t, es-tis, differ from leg-ĭ-t, leg-ĭ-tis merely in the absence of the connecting vowel ĕ (class. Lat. ĭ, ch. ii. § 14). So ī-s (older *ei-s), ī-tis (older *ei-tis) from the root ei-, ‘to go’; vul-t (older vol-t), vul-tis (older vol-tis) from the root wel-, ‘to wish’ (ch. ii. § 8; on vīs see § 23). The -m of sum is the Athematic -mī of the 1 Sg. Pres. Ind.; and the īē (ī) of the Optative appears in the so-called ‘Subjunctive,’ s-īē-s (class. sīs), s-ī-tis. But in Latin a great many thematic forms are mixed with these Athematic Verbs. Thus their First Singular Pres. Ind. is usually thematic, e-ō from ey-ō (ch. x. § 13), vol-ō, and their Third Plural of the same tense, e.g. eunt from ey-ont (ch. ii. § 14). And throughout the Latin Verb we shall find a strange mixture of thematic with athematic forms.

To this Conjugation belongs the large number of Derivative Verbs from Ā-stem Nouns, e.g. from planta, a plant, plantare, to plant, ‘to make into a plant,’ from fuga, flight, fugare, to put to flight. This form of Verb became associated with a Transitive sense, ‘to make,’ ‘reduce anything to a certain state or condition,’ and was extended to Verbs derived from other stems, if these Verbs had this transitive meaning, e.g. colorare, to make coloured, from color, scelerare, to make guilty, from scelus, pulverare, to turn into dust, to cover with dust, from pulvis. A section of the First Conjugation Verbs belongs to a very ancient type, such as stā-re from the root stā-, ‘to stand,’ in-trā-re, where trā- is a Verb-root, a derivative of the root ter-, ‘to go through,’ ‘pierce.’ This ancient type belonged properly to the Athematic Conjugation, while Derivatives like fugare were thematic and used the suffix -yŏ- to form their Present Tense. The two types have been blended in Latin in such a way that both form their 1 Sg. Pres. Ind. thematically, with the addition of this suffix -yŏ-, and the rest of their Persons athematically; and so we have: intrō (from -trā-yŏ) like fugō (from -gā-yŏ); fugā-s like intrā-s, fugā-tis like intrā-tis. [1]

[^1]: The athematic form would be i-nt, a form which seems actually to have been in use.

The First Conjugation was the favourite Latin Conjugation, and verbs taken from other languages were impressed into it, [2] e.g. propīnare, to drink a person’s health, from Greek προπίνειν, atticissare, to ape the Athenian fashion, from Greek ἀττικίζειν. It includes the numerous class of Iteratives or Frequentatives, e.g. pulso, I strike frequently, the Frequentative of pello, formed from Perf. Part. Pass.-stems or rather from the Fem. of these used as a Noun (cf. offensa beside offensus, repulsa beside repulsus). They sometimes double the TO-suffix, e.g. fac-tĭ-to, ven-tĭ-to.

[2]: So in German loan-verbs take the ending -ieren, e.g. commandieren.

O. Lat. sonĕre, e.g. Lucr. iii. 156:

Caligare oculos, sonere auris, succidere artus,

comes directly from the root swĕn- (Lat. sŏn-, ch. x. § 13) as legere from the root leg- (see § 4), and formed its Perfect and Supine regularly, sonui, sonitum. Sonāre, a Derivative from a lost Verbal Noun sona, existed side by side with sonere and in time ousted the Third Conj. form, though the Third Conj. Perf. and Supine were retained. Hence the irregular conjugation: sono, sonui, sonitum, sonare. Similarly the coexistence of im-plĭco (3 Conj.) from the root plek-, ‘to fold’ (cf. Gk. πλέκω) and im-plīco (1 Conj.) from a lost Verbal Noun of the First Decl. (cf. Gk. πλοκή) gave rise to the blended conjugation: implico, implicui and implicavi, implicitum and implicatum, implicare. In the case of cēlo from a lost Verbal Noun cēla, concealment, from the root cēl-, the Third Conj. form cĕlo is found only in Compounds, e.g. oc-culo (from *ob and cĕlo); while in profligare beside fligere, occupare beside capere, &c., it is the First Conj. form which appears in the Compounds, and the Third Conj. form in the Simple Verb. (On dare, with dă-, the weak grade of the root dō-, ‘to give,’ and on stāre from the root stā-, ‘to stand,’ see § 23.)

Here too we may discriminate an ancient type of Athematic Verbs, e.g. im-plē-re with the root plē- (cf. P.P.P. im-plē-tus), a by-form of the root pel-, ‘to fill,’ from the great mass of Derivative Verbs, e.g. claudē-re, to be lame, derived from the O-stem Adjective claudus, lame. These Derivatives of the Second Conjugation contrast with the Derivatives of the First in that they are derived not from Ā-stems but from O-stems and have not a transitive but an intransitive meaning. Clarere, for example, means ‘to be clear,’ while clarare means ‘to make clear.’ Like the Derivative Verbs of the First Conjugation these Derivatives of the Second formed their Present Tense stem originally with the suffix -yŏ- and belonged properly to the Thematic Conjugation. In Latin the intermixture of the thematic and athematic forms has had the same result as in the First Conjugation, namely that the First Sing. Pres. Ind. was formed thematically with the YO-suffix, the other Persons athematically, e.g. impleo (from -ē-yŏ) like claudeo; claudē-s, claudē-tis, like implē-s, implē-tis.

But there is a further element of confusion in the Second Conjugation. The I.-Eur. Causative and Intensive Verbs, whose Present-stem was formed in -ĕyŏ- with the O-grade of the root (ch. x. § 12), were attracted into this Conjugation in Latin; and so, e.g. moneo, I remind, cause to remember (with Present-stem mŏnĕyŏ-), the Causative of the Verb-root men-, ‘to remember,’ is a Second Conjugation Verb. Another verb of this form with Causative meaning is torreo, I cause to become dry, from the root ters-, ‘to be dry.’ Another with Causative or Intensive meaning is noceo from the root nec- (Lat. nex). The absorption of these verbs into the Second Conjugation brought about the strange result that though the Second Conjugation type is associated with an intransitive sense, a small body of verbs of this Conjugation have a decidedly transitive meaning.

The intransitive character of these Second Conj. Verbs, the Ē-Verbs as they may be called, is due to the fact that the Verb-stem with this suffix -ē- had an intransitive sense. In Greek the Verb-stem with the Ē-suffix was used as a Passive, e.g. ἐ-τύπη-ν, I was struck, or merely an Intransitive, e.g. ἐ-μάνη-ν, I was in a state of madness. In Latin this intransitive Verb-stem in ē appears in cale-facio, to put into a state of warmth, cale-fio, to be put into a state of warmth; closely connected with it are Verbal Noun-stems in ē, e.g. sordēs Plur. (cf. sordē-facio), the declension of which varies between the Fifth and the Third Declension (see ch. iii. § 13). There are a large number of these Verbs in -eo indicating state or condition in Latin, e.g. caleo, timeo, paveo, each with a corresponding Noun in -or (an S-stem, ch. iii. § 8; ch. xi. § 18), e.g. calor, timor, pavor, and an Adj. in -ĭ-dus, e.g. calidus, timidus, pavidus. ‘Inceptive’ Verbs, which denote the passing into a state or condition, e.g. calesco, to pass into the state of warmth, liquesco, to pass into a liquid state, are so closely associated with Intransitive Ē-Verbs that they are often used as their Present Tense, e.g. inardesco (inardeo is not found), convalesco (not convaleo). In fact it is a law of Latin that in composition with the Prepositions cum, ex, in, Prepositions which convey the idea of ‘becoming,’ these Ē-Verbs must form their Present Tense in -esco, unless the Preposition retains its separate force, e.g. e-luceo, to shine out (not ‘to pass into a shining state’), co-haereo, to be united with (not ‘to pass into a united state’). So remote are these Verbs from a transitive sense that a construction like horrere aliquem, pavere aliquem is not found before Cicero’s time. To make Transitives of them the Auxiliary Verb facio was brought into requisition, e.g. pavē-facio, calē-facio.

Third Conj. Verbs which had an intransitive meaning tended to pass into the Second Conjugation. Thus fervo, the older form, e.g. Virg. Aen. iv. 567:

Jam fervēre litora flammis,

became in class. Lat. ferveo. (Cf. fulgeo from fulgo, scateo from scato, intueor from intuor.)

This is the common receptacle for Verbs of all kinds but the Derivative YŎ-Verbs, derived from Nouns or Adjectives, which belong to the First, Second, and Fourth. Such Derivatives from U-stems however are included in this Conjugation, e.g. statuō from *statuyo, derived from status. Verbs in -io not derived from Nouns or Adjectives, which for a time wavered between the Third and Fourth Conjugations, have been mostly absorbed by the Third, such as facio (2 Sg. facis 3 Conj., but older facīs 4 Conj.), pario (Inf. parĕre, but in Plautus parīre). The Derivatives from Nouns and Adjectives are distinguishable from them by their Perfect in -īvi and P.P.P. in -ītus, e.g. finivi, finitus, though even this form of the Perf. Ind. and Part. Pass. occurs in a few Third Conj. Verbs like lacesso, which had by-forms in -io, *lacessio, *capessio. The types of Verb which it contains are therefore very numerous. We have (1) the common type of Thematic Verb, with the ordinary form of the root, e.g. leg-ō from the root leg-, dīco (older deico) from the root deic-, dūco (older douco with ou for I.-Eur. eu, ch. x. § 11) from the root deuc-; (2) with Reduplicated Root, e.g. gi-gn-o (Gk. γί-γν-ομαι) from the root gen-; (3) with Nasalized Root, the nasal being either inserted in the root, e.g. ju-n-g-o from the root jug- (I.-Eur. yeng-, weak form yŭg-), to join, li-n-quo from the root liq-, leiq-, to leave, or appended, e.g. ster-n-o from the root ster-, to strew, li-n-o from the root li-, lei-, to smear. Sometimes a syllable is appended, e.g. sternuo, I sneeze, minuo, a type of Verb which corresponds to Greek Verbs in -νῡμι or -νῡω, e.g. πτάρ-νῡ-μαι (cf. μι-νύ-θω). (4) with YŎ-suffix, e.g. specio. [3] These wavered between the Fourth and Third Conjugations, unlike the Derivatives from I-stems with 1 Sg. Pres. in -io from *-iyŏ, e.g. finio, which belong exclusively to the Fourth (see below, § 5). (5) ‘Inceptives’ with the SKŎ-suffix, which are not rightly called ‘Inceptives,’ since they denote the passing into a state or condition, ‘becoming’ rather than ‘beginning,’ e.g. calesco, I become warm, liquesco, I pass into a liquid state (§ 3), and the archaic ‘Inceptive’ of sum, esco (used by Lucretius). These modifications of the Root by Reduplication, Nasalization, addition of YŎ- and SKŎ-suffixes were made with the object of forming a Present-Tense Stem, and were properly discarded in the Perfect and the tenses related to the Perfect. The stem gigno- of gigno had the idea of ‘continued production,’ so that gigno meant ‘I continue to produce,’ ‘I am in the state of producing.’ To form the Tenses expressing momentary, past action such a stem could not be used; so we find the Perfect genui exhibiting the root in the form gen-, not in the form gign- (cf. Gk. ἐ-γεν-ό-μην 2 Aor., beside γίγν-ο-μαι Pres.). These Tense-stems will be discussed afterwards (§§ 7–12); but meanwhile we may point out that by a not unnatural confusion a stem which properly belonged only to the Present Tense was often retained throughout the Verb (e.g. junxi, junctus retain the n of the Pres.-stem), or a stem was used in the Present which properly belonged to other Tenses, e.g. rūlo, beside rūdo (older *reudo), geno Lucr. (On the use of ‘Inceptives’ in -sco as Present of Intransitive Ē-verbs, e.g. incalesco, not *incaleo, see above, § 3.)

[3]: Y in Greek after a consonant became τ, so Gk. σκέπτομαι, for *σκεπ-yo-μαι, corresponds to Lat. specio.

This like the First and Second, consists mainly of Derivative Verbs. These are not merely Derivatives from I-stems, e.g. finio (Present-stem fini-yŏ-) derived from the I-stem finis, but Derivatives of various origin, all showing the usual Derivative Present-stem suffix -yŏ-, e.g. custodio, I am a guard, from custos. As the First Conj. Derivatives had a transitive, the Second Conj. Derivatives an intransitive sense, so the Fourth Conj. Derivatives are to some extent associated with the idea of a mental or bodily state, especially a state of disease, e.g. rancio, I am hoarse, dentio, I am teething. Desiderative Verbs (which were used more in conversational Latin than in the literary language) belong to the Fourth Conjugation, e.g. esurio, I am hungry, I desire to eat, empturio, I desire to buy. They are formed with the YŎ-suffix from Verbal Noun-stems in -tor-, e.g. esurio (older *esorio) from esor (ed-tor-), an eater, empturio (older *emptorio) from emptor, a buyer, and must be distinguished from a small class of Verbs in -urio (also written -urrio), e.g. ligurrio, I lick, scaturio, I gush, which are Derivatives from Verbal Nouns in -uris, e.g. securis, an axe (from seco), or -ūra, e.g. figura (from fingo).

The Fourth Conjugation is connected with the Third by the accident that a large number of Verbs of the Third Conjugation had by-forms made with the YŎ-suffix, e.g. pinsio beside pinso (cf. lacessivi, &c., from *lacessio, a by-form of lacesso), and that several of the YŎ-verbs did not retain this suffix throughout their Conjugation, e.g. facio has Inf. facĕre, cupio has Inf. cupĕre. Facio and cupio are by the classical period entirely separated from any connexion with the Fourth Conj., though Plautus has facīs, cupīs like audīs, and a hexameter line of Ennius ends with corde cupītus; and the same is true of adgredior (but adgredīmur Plaut.), adorior (but adorītur Lucr.), morior (but morīmur Enn.), and the like. It is perhaps to the connecting-link furnished by these Verbs in -io of the Third Conjugation that we should refer the remodelling of the Imperf. and Fut. Ind. of the Fourth Conj. on the type of the Third, which was fully effected before the classical period; finiebam replacing the older finibam, finiam the older finibo (see §§ 8, 9).

The three Voices of Greek, Active, Middle, and Passive, are in Latin Grammars reduced to two, Active and Passive. The Latin Deponents however take to some extent the place of the Greek Middle, and sequor is an example of an I.-Eur. Middle which appears in the I.-Eur. languages with the Middle or Passive endings (e.g. Gk. ἕπο-μαι, Early Irish sĕchur). These endings are explained in § 21. In the Italic and Celtic languages their distinguishing feature is the letter r (e.g. Lat. sequor, sequitur; O. Ir. sĕchur, sĕchĕthar), and there are traces that the passive R-forms were originally restricted to an Impersonal use in which this r was added immediately to the root of the verb. If this be so, a supposed form ama-r would in very early times mean ‘there is loving’ and would govern an Accusative Case; later it would receive a Personal suffix, *ama-tŏ-r (class. Lat. amatur, ch. ii. § 14), and gradually take the Noun as a Subject not as an Object. Amar hominem would become amatur hominem, then amatur homo. This Impersonal use of the Passive was a great feature of Latin, e.g. itur in antiquam silvam, Virg.; PRIVATVM. PRECARIO ADEITVR, ‘No Admittance,’ on an inscription. (On the Fut. Inf. Pass. amatum iri see § 15.) In the early literature we find occasionally the Noun in the Accusative, e.g. vitam vivitur Enn.

Another trace of the I.-Eur. Middle in Latin is the Perfect Active, whose ending -ī (older -ei) shows it to be a Middle and not an Active form (see § 10), so that a Perfect like reverti goes naturally with a present revertor. On the other hand the earlier Perfects gavisi, [4] solui, ausi were in classical Latin replaced by the usual Deponent type of Perfect, gavisus sum, solitus sum, ausus sum; and, by a curious attraction, coeptus sum, desitus sum were used when a Pass. Inf. followed, e.g. urbs coepta est aedificari. [5]

[4]: E.g. Liv. Andronicus: quoniam audiui paucis gauisi.

[5]: So in the early writers potestur, poteratur, possetur, nequitur, nequitum, e.g. Lucr. iii. 1010: quod tamen expleri nulla ratione potestur.

The I.-Eur. Middle was associated with Verbs indicating states of feeling, operations of the senses, condition of life, &c.; and the Latin Deponents are still true to this type, e.g. reor, vereor, irascor, contemplor, poetor, dominor, auguror, aemulor. [6] Like the Reflexive use of the Greek Middle is Lat. cingor for cingo me, &c. Examples of Frequentative Middles are hortor (Frequentative of O. Lat. horior), meditor (from a lost medor, Gk. μέδομαι), imitor (from a lost imor; cf. imā-go), nitor for *nivitor (from a root with a Guttural; cf. nixus, and see ch. x. § 18).

[6]: Lat. -ari corresponds to Greek -εύειν, e.g. Lat. dominari to Greek τυραννεύειν, ‘to be a tyrant,’ ‘to play the part of a tyrant.’

As in Greek ἀποθνῄσκω — a Verb of Active form — is used as the Passive of ἀποκτείνω, so in Latin vapulo is the Passive of caedo, to beat, vēneo (venum eo) of vendo, to sell, fio of facio, to make. The older Inf. fiere [7] was in classical Latin changed to suit the form of the ordinary Passive Inf. and became fieri, much as Plautus uses vēniri for vēnire.

[7]: Ennius in the description of Homer’s Metempsychosis has: memini me fiere pauom, ‘I remember becoming a peacock.’

The various Tenses of the Verb had each their Tense-stem, formed from the root of the Verb by the addition of some suffix or by some other modification (e.g. Reduplication, § 10). In addition, the Past Tenses might be distinguished by the Prefix of the Augment e-, a Demonstrative Particle signifying ‘then,’ ‘there’; but in the earliest Greek and Sanscrit literature this Particle is often omitted and in Latin there seems to be no trace of it. The formation of the Present-stem has already been treated in the account of the Latin Conjugations. To indicate the idea of present or continued Action we have seen that the root of the Verb was modified in the First, Second, and Fourth Conjugations by the addition of a YO-suffix; e.g. amā-yŏ-, is the Present-stem of the Verb amare, to love, whence 1 Sg. Pres. Ind. *amā-yŏ, which became amo; so pleo for *plē-yŏ (stem plē-yŏ-), &c. Third Conjugation Verbs show various modes of forming the Present-stem, such as (1) Reduplication, e.g. gigno (stem gi-gnŏ-; cf. Gk. γίγνομαι) from the Root gen-, (2) N-suffix, e.g. ster-n-o from the root ster-, li-n-o from the root lei-, or N-infix, e.g. pă-n-g-o from the root pag-, this n disappearing usually in the Perfect, &c., e.g. strā-vi, lī-vi, pē-pig-i or pēg-i. A common Third Conjugation type however shows the ordinary form of the root with no addition but the thematic vowel (e.g. leg-o from the root leg-; dīco, originally deic-o, from the root deic-). The SKO-suffix gave the notion of becoming, and so was adapted to the Present Tense, e.g. incalesco, I become warm, Perf. incalui (see § 3).

This Tense belongs to the Present-system, being really the Preterite of the Present Tenses. Its Preterite sense is given to it in Latin by the addition in the First Pers. Sing. of -bam. The -bam of this Tense is really an Auxiliary Verb, a Preterite of the I.-Eur. root bheu-, ‘to be’ (Lat. fui, O. Lat. fuo; cf. ch. x. § 16), appended to a Verb-stem, e.g. amā-bam, vidē-bam, legē-bam, audī-bam (class. audie-bam, remodelled after the Third Conjugation, § 5). Eram stands alone in dispensing with this Auxiliary. It is a Preterite from the root es-, ‘to be,’ of the same formation as -bam from the root bheu-.

Like the -bam of the Imperfect, the -bō of the Future conceals an Auxiliary Verb, a Future, or rather Subjunctive of the root bheu-, ‘to be.’ Erō, which corresponds to the Homeric ἔω, Att. ὦ (from *ἔσω), is a corresponding Subjunctive from the root es-, ‘to be,’ with that future sense which belonged to the Subjunctive in early times (p. 100 n.). This formation of the Future in -bo is peculiar to the three Conjugations, which are especially the Conjugations of Derivative Verbs, that is to say to the First, Second, and Fourth, e.g. amā-bo, vidē-bo, and in Early Latin audī-bo, though by the classical period the Fourth Conj. Future was remodelled on the pattern of the Third Conjugation, audiam like legam (§ 5). This so-called Future of the Third and Fourth Conjugations is nothing but the Subjunctive (§ 13), the Ā-Subjunctive form being used in the First Person Singular, the Ē-Subjunctive forms in the other Persons, e.g. legam (originally legā-m, ch. ii. § 6), legē-s, legĕ-t (originally legē-t, ch. ii. § 16). (On these Subjunctive forms see § 13.)

The completeness with which the Verb-system has been preserved in Greek in contrast with Latin is nowhere more clearly seen than in the Preterite Tenses. While Greek preserves separately (1) a First Aorist, better called the S-Aorist, e.g. ἔ-δειξ-α for *ἐ-δεικ-σ-α, from the root δεικ-, (2) a Second Aorist, better called the Weak-root Aorist, because it shows the weak root (ch. x. § 12) of the verb, e.g. ἔ-πιθ-ον from πιθ-, the weak form of the root πειθ-, I.-Eur. bheidh-, (3) a Perfect, showing the Reduplicated Verb-root, e.g. μέμονα, we find in Latin all three mixed up together into a single Preterite, which we call the Perfect Tense, and all driven into the same groove of declension. The S-Perfect, which corresponds to the Greek First Aorist, e.g. dixi, earlier deix-ei (like ἔ-δειξ-α), the true Perfect, e.g. memin-i (like Gk. μέμον-α), and Perfects like fidi, scidi, which some compare with the Greek Second Aorist, are all declined in the same way: dixi, dixisti, dixit; memini, meministi, meminit; scidi, scidisti, scidit. On the other hand we have in Latin a type of Preterite which is not found in Greek, the V-Perfect, e.g. amā-v-ī, mon-u-ī, audī-v-ī, the origin of which is not known.

The declension too of the Perfect shows traces of having been pieced together from various materials. The ending ī (older ei) of the First Person Singular is I.-Eur. -ai, the ending of the Middle Voice. The Third Sing. of the Perfect Middle had the same ending in I.-Eur., but in Latin -t (the usual 3 Sg. suffix, § 20) has been added to distinguish the Third from the First Person, -ei-t becoming -īt, then -ĭt (ch. ii. § 16). The Third Person Plural in -runt, older -ront (ch. ii. § 14), shows the same r that appears often in the 3 Pl. of the I.-Eur. Verb. The First Person Plural is an Active form, e.g. meminimus (cf. Gk. μεμόναμεν).

Although the I.-Eur. Perfect had as a rule a Reduplicated stem, there were a few Unreduplicated Perfects, e.g. woid-, the Perfect-stem from the root weid-, ‘to see, know’ (Greek οἶδα represents the Active, Lat. vīdī the Middle), sēd-, the Perfect-stem from the root sĕd-, ‘to sit’ (Lat. sēdī). In the second example the raising of the root-vowel from ĕ to ē seems to take the place of Reduplication, parallel to which we find in Latin pēgi (beside pēpigi), lēgi (root lĕg-), ēgi (root ag-) and the like. But undoubtedly the original Perfect-type had Reduplication, and most apparently Unreduplicated Perfects in Latin like tŭli, scĭdi were Reduplicated at an earlier period of the language, [8] and lost their Reduplication first in Compounds under the Early Accent Law (ch. ii. § 12), then by analogy in the Simple Verb also. Re-tetuli became ret(ĕ)tuli (hence always spelt with double t) with Syncope of ĕ, at-tetuli became attuli, and so on, until at last the use of -tuli in Compounds led to the use of tuli in the Simple Verb. Similarly we have cucurri but concurri, incurri, etc.

[8]: Tetuli is the only form known to Plautus; cf. O. Lat. scicidi.

The Vowel of the Reduplication Syllable was originally ĕ, but owing to the Latin tendency to Assimilation (ch. ii. § 15) adapted itself to the vowel of the following syllable, e.g. cucurri, momordi. This was the formation in vogue at the classical period, but in the earlier literature we find cecurri, memordi, &c.; and even Cicero and Caesar seem to have allowed the older forms.

The shortened forms of the V- and S-Perfects, e.g. audisti for audivisti, dixti for dixisti, are produced by two tendencies of the language, (1) the tendency to drop v between two vowels, especially between two similar vowels, such as i-i (cf. obliscor an old by-form of obliviscor, I forget, sīs for sī vīs, ‘if you please,’ dīnus, an old by-form of dīvīnus) (ch. ii. § 12), (2) the tendency to drop one of two similar neighbouring syllables (cf. semōdius for *semi-modius, idolatria for *idolo-latria, &c., ch. ii. § 12). We have dixti for dix-is-ti, misti for mi-si-sti, where the two syllables are similar, but not, e.g. *cepsti for ce-pi-sti, where the two syllables have not similarity of sound.

Like the Imperfect in -bam, and Future in -bo, the Pluperfect in -eram is an Auxiliary formation, -eram being the Preterite of the root es-, ‘to be’ (§ 8), which is added to the Perfect-stem, e.g. amāv-eram, monu-eram, dīx-eram, memin-eram, audīv-eram.

This Tense adds to the Perfect-stem the Auxiliary ero (§ 9), as the Pluperfect adds eram, e.g. amāv-ero, monu-ero, dīx-ero, memin-ero, audīv-ero. In Early Latin there were forms in -sso, which are replaced in classical Latin by Fut.-Perf. forms, but which are of an entirely different origin, e.g. amasso. They seem to have been originally Subjunctives of some S-tense, like the Greek S-Aorist, and are in fact the exact equivalents of Greek Futures like τιμήσω, which stands for τιμάσω, and is properly the Subjunctive [9] of ἐτίμησα. Similarly dixo (cf. faxo) may be compared with Gk. δείξω. (On the Optative of this Latin S-Aorist, amassim, etc., see § 13.)

[9]: I.e. the Athematic Subjunctive, which had ο, ε where the Thematic had ω, η. Thus τιμήσομεν (cf. τείσομεν Subj. in Homer) is Athematic, τιμήσωμεν Thematic. The I.-Eur. Subj. had often a Fut. sense, a feature of the old language which is still retained in the Greek of Homer, e.g. Od. xvi. 437 οὐδ’ ἔσσεται οὐδὲ γένηται.

As the Latin Perfect Tense combines Perfect- and Aorist-forms (§ 10), so the Latin Subjunctive Mood includes the relics of the Optative along with the genuine Subjunctive forms. Optative are all the so-called ‘Subjunctives’ in -im. They show the Athematic Optative, which properly had -iē- in the Singular, -ī- in the Plural (§ 1), a type preserved only in Old Latin siem, sies, siet, sīmus, sītis (sient); and even that has by the classical period been reduced like the rest to the uniform use of the ī-forms, sim, sis, sit. So edim, edis, edit; velim, velis, velit, and in the older language duim, duis, duit.

Of the two Subjunctive-types, the Ā-Subjunctive and the Ē-Subjunctive, the former is used by the Second, Third, and Fourth Conjugations, e.g. cale-am, cale-ās, cale-at; leg-am, leg-ās, legat; audi-am, audi-ās, audi-at, the latter by the First, e.g. am-em, am-ēs, am-et. The reason for the use of the Ē-type in the First Conjugation is that the Ā-type must have been confused with Pres. Indicative forms, e.g. amās, amat. The Ē-type was also pressed into the service of the Third and after its model (§ 5) of the Fourth Conjugation in the capacity of a Future Indicative (§ 12 n.), this type being chosen for the sake of distinction from the Pres. Subjunctive; though curiously enough in the First Person Singular of the Future, the Ā-form was used, e.g. legam, leges, leget; audiam, audies, audiet. The reason for the retention of the Ā-form in this Person was that this single Person had been in use as a Future long before the fashion was introduced of dropping the proper Future-tense (§ 9) and using Subjunctive-forms in its place. Audiam had established itself in use before audies, audiet replaced audibis, audibit, &c., and though we find traces of a temporary effort at uniformity by the substitution of audiem, faciem, &c., the old-established form maintained its ground.

The Ē-type was adopted by the true Subjunctive of sum, if we may infer this from the fact that sem and not sam is the Auxiliary used in the formation of the Imperfect and Pluperfect Subjunctive, e.g. amā-rem, amā-res, amā-ret, with r from an older s (ch. x. § 19); amavis-sem, amavis-ses, amavis-set. Vellem comes from *vel-sem (ch. x. § 19), ferrem from *fer-sem (ch. x. § 19). But from fuo, an obsolete Pres. of which fui is the Perfect, we have fuam in Old Latin; cf. Virg. Aen. x. 108:

Tros Rutulusve fuat nullo discrimine habebo.

Forem seems to stand for *fuerem, as Marcipor, an Old Latin slave-name, for *Marcipuer. Edim and edam, duim and duam are the last examples of the coexistence of rival Optative and Subjunctive forms in Latin. They illustrate how, the Opt. and Subj. having become identical in meaning (both edim and edam have the meanings ‘Oh that I may eat,’ ‘let me eat,’ ‘[that] I may eat’), one of the forms (usually the Optative) was dropped as superfluous.

The old Optative sense is preserved in the ante-classical Optative-forms in -ssim, e.g. amassim, which occur in early Latin prayers, e.g. Juppiter prohibessis scelus; di mactassint, and the Augural formula: bene sponsis beneque uolueris. These forms in -ssim, often called Perfect Subjunctives, are the Optative Mood of the forms in -sso, e.g. amasso, mentioned in § 10; they are better called Precatives.

In the 2 Sg. Act. of the Present Imperative the bare stem of the Verb was used, e.g. Athematic ei (class. Lat. ī) from the root ei-, to go, Thematic lege from the root leg-, to gather. But a particle was often added, e.g. -θι in Gk. ἴ-θι, &c.; and so -tōd (which became -tō, ch. x. § 17) (Abl. Sing. of the Demonstr. Pronoun stem to-, ch. v. § 3) to the 2, 3 Sg., e.g. es-tō, older estōd (Gk. ἔσ-τω). The addition of -tōd ‘from that,’ ‘thereupon,’ to the 2 Sg. Pres. Imperat. gives it a Future Imperat. sense, e.g. Hor. C. iii. 14. 23:

Si per invisum mora janitorem Fiet, abito,

‘if the surly porter detain you, go away,’ lit. ‘go away thereupon.’ Memento is I.-Eur. *me-mṇ-tōd (Gk. μεμάτω). The 2 Plur. of the Future Imperative adds -tĕ (the 2 Plur. suffix, e.g. abi-te) to the Sing. form, e.g. abito-te.

Fer, vel (used as a Conjunction, ch. ix. § 3), es are regularly formed Athematic Imperatives from the roots fer- (I.-Eur. bher-), vel- (I.-Eur. wel-), es-. Another is -do (originally -dō, ch. ii. § 16) of cĕ-dō, ‘give here,’ from cĕ-, the Demonstr. Particle (ch. v. § 4) and the root dō-, ‘to give’ (cf. Gk. δί-δω, Imperat. of δί-δω-μι). The genuine Imperat. of dare has been preserved in this Compound only, for dā follows the Analogy of the First Declension. But dic, duc, fac are Thematic Imperatives, representing an earlier dice, duce, face (ch. ii. § 12), and so is em (used as an Interjection, ch. ix. § 18), properly the Imperat. of emo, I take. (Cf. Catullus xxvii. 2: inger mi calices amariores, instead of ‘ingere mi.’)

The I.-Eur. Infinitive was merely a Case (usually Dat. or Loc. Sg.) of a Verbal Noun. Thus Lat. da-rī (O. Lat. da-sei) is Dat. of an S-stem like the Noun generi, Dat. of the S-stem genus (ch. xi. § 18); Lat. da-rĕ (earlier da-sĕ) is Loc. of the same stem (on generĕ, Loc. used as Abl., see ch. iii. § 8); similarly es-sĕ from the root es-, to be, dedis-sĕ, fer-rĕ for *fer-sĕ (ch. x. § 19), vel-lĕ for *vel-sĕ (ibid.). For Inf. Pass. of the Third Conjugation we have the Dat. not of an S-stem but of a Root-stem (ch. xi. § 19), e.g. leg-ī (not leger-ī). Thus the Latin Inf. Pass. differed from the Inf. Act. only conventionally, the Dat. case being reserved for the one use, the Loc. for the other, and had no distinctive Passive suffix. Whether this is present in the O. Lat. by-forms legier, darier, etc., is not certain.

For the Perfect Inf. Passive the Perf. Part. Pass. was used with the Auxiliary Verb esse, e.g. constat id factum esse, constat ea facta esse; for the Fut. Pass. the 1st Supine with iri, Inf. Pass. of eo, to go, e.g. constat id factum iri, constat ea factum iri. [10] The Fut. Act., e.g. constat id eventurum (esse), is most naturally explained as a combination of the Fut. Part. Act. with esse; though its Old Latin indeclinable use, e.g. credo inimicos meos dicturum (from a speech of C. Gracchus), has suggested the theory that it is a compound of the 2nd Supine in -tū with a supposed old Inf. of sum, viz. *erum (from *esŏm), dicturum for *dictū-erum being in time made personal dicturus, -a, -um (the Fut. Part. Act.) in the same way as O. Lat. ‘dicendum est orationem’ changed to class. Lat. ‘dicenda est oratio’ (§ 18).

[10]: Iri is Impersonal Passive like Virgil’s itur in antiquam silvam (§ 6), so that the sentence literally means ‘it is agreed that there is a going to do these things.’ A quotation from a speech of Cato: contumelia mihi factum itur, shows the same tendency to make this Impersonal Passive personal as produced vita vivitur out of O. Lat. vitam vivitur (§ 6).

The First Supine, used after a Verb of motion, is the Acc. Sg. of a Verbal Noun, a TU-stem (ch. xi. § 13), e.g. ire spectatum, lit. ‘to go to the seeing,’ like ire domum, to go to the house, ire Romam, to go to Rome.

The Second Supine, used after an Adjective, is the Loc. Sg. (ch. iii. § 11) of the same Verbal Noun, e.g. agilis cursu, nimble in running. This Loc. Sg. in -ū of U-stems often played the part of a Dat. (cf. curru for currui in Virgil); and we find the Second Supine used not only as a Locative, but as a Dative, e.g. (fabula) lepida memoratu, pleasant for telling, where in the older language the Dative proper in -ui is used, e.g. lepida memoratui (Plaut.), as well as the Locative, e.g. ridicula auditu (Plaut.). [11]

[11]: In Plaut. Rud. 294 this Loc. plays the part of a Predicative Dative: Sunt nobis quaestu et cultu, ‘They are our trade and pursuit.’

This TU-stem bulks largely in the language of Plautus and the older Dramatists, e.g. obsonatu redeo, I return from buying food for dinner, essum vocare, to invite to dinner, nuptum dare, to give in marriage. In Aul. 736 perditum ire is used almost like perdere:

Quamōbrem ita faceres mēque meosque perditum ires liberos,

(cf. Bacch. 565 mi ires consultum male, ‘you would go and thwart my interests’). The use of the Accusative without a Preposition is common in early Latin in phrases like i malam crucem, go and be hanged (Plaut.), suppetias ire, to go to help, infitias ire, to deny, exsequias ire, to go to a funeral. The Supine use thus arose naturally out of the tendencies of the language. Like nuptum dare and nuptum ire are vēnumdare or vēnundare (vendere) and vēnum ire (vēnire); pessumdare or pessum dare and pessum ire, which may be regarded as survivals of the Plautine idiom.

The I.-Eur. Participles were merely Verbal Adjectives formed with the various suffixes mentioned in ch. xi. Thus for the Perf. Part. Pass. the TO-suffix was used to form certain Verbal Adjectives which in Latin took the function of Perfect Participles Passive (Gk. θε-τός, Lat. crē-di-tus, re-plē-tus), [12] or the NO-suffix (Engl. bound-en; cf. Lat. plē-nus Adj.). For the Gerundive the YO-suffix was used in various I.-Eur. languages, traces in Greek being words like ἅγ-ιος, venerable, ‘worthy of veneration’ from ἅζομαι, and in Latin, eximius, ‘worthy of being taken out,’ from exemo, to take out.

[12]: Aegrōtus may be a trace of a Latin O-Conjugation like Greek δηλόω.

The Pres. Part. Act. (and all Active Participles, except the Perfect), [13] took the suffix -ĕnt- (-ŏnt-, -nt-; ch. xi. § 12), e.g. Gk. φέρ-ων, -οντος, Lat. fer-ens, -entis. The Pres. Part. Middle (or Passive) was formed in -mĕnŏ- (-mono-, -mno-; ch. xi. § 12), e.g. Gk. φερό-μενος, but dropped at an early period out of use in Latin; though it is found in the 2 Pl. Ind. Pass., e.g. ferimini (sc. estis; see § 21), and in nouns like alumnus (cf. Gk. ὁ τρεφόμενος, see ch. xi. § 6). (On 2 Pl. Imper. ferimini, see § 21.) The Fut. Part. Act. in -tūrus is probably a formation with the suffix -rŏ- from a TU-stem Verbal Noun, e.g. scripturus (stem scripturo-) from the stem scriptu- of scriptus, -ūs, picturus from the stem pictu- of pictus, -ūs, &c., like Gk. ἰσχυ-ρός from ἰσχύς. (For another explanation, see § 15.) For a Participle the Latin writers, especially the poets, often substituted an Adjective, e.g. lacer for laceratus (lacerum crudeliter ora Virg.), and these Adjectives or ‘truncated Participles’ have to some extent encroached on the Perf. Part. Pass. in the Romance languages, e.g. Ital. trovo beside trovato, ‘found.’ (On Verbal Adjectives in -bilis, see ch. xi. § 12.)

[13]: Its suffix was -wĕs-, -us-. Some find traces of a Perf. Part. Act. in Latin in words like cadā-ver, papā-ver; others in O. Lat. gnārūres Plur., knowing (e.g. Plaut. Most. 100). Mentor is not a Perf. Part., but an Adj. derived from a Perf. Part. stem, as Gk. κεκραγμός (Eurip.) is a Noun derived from κέκραγα.

The close relation between Participles, Adjectives, and even Nouns is seen in words like rudens, a rope, lit. ‘rattling’; benevolens, a friend, used as a Noun by Plautus. [14] It was this close relationship of benevolens and benevolus, insciens and inscius, indigens and indigus, congruens and congruus, &c., which led to a type of Comparison like benevolus, -entior, -entissimus (ch. iv. § 3). Very early examples of Pres. Participles used as Nouns or Adjectives are dens, a Pres. Part. of the root ed-, to eat, and sons, a Pres. Part. of the root es-, to be, so that dens properly means ‘the eater,’ sons, [15] ‘being,’ ‘truly being,’ whence ‘truly charged,’ ‘guilty.’ (Our ‘sooth’ is the same word.) The Perf. Part. Pass. had the same tendency to become an Adjective, e.g. citus, swift, lit. ‘bestirred’; catus, originally sharp, [16] from the root cō-, to sharpen (cf. cōs, a whetstone), then (metaphorically) (1) piercing, of sounds, e.g. cata signa, Enn., ‘the shrill clarions,’ (2) shrewd, of persons; lātus, O. Lat. stlātus (ch. x. § 19), broad, lit. ‘extended,’ from the root stel-, to spread, extend. Of Past Parts. Pass. as Nouns we have e.g. natus, a son, legatus, a lieutenant, deputy. When used as Adjectives, they sometimes passed into I-stems, the favourite form of stem for Adjectives (ch. xi. § 12), e.g. fortis, O. Lat. forctus, originally P.P.P. of the I.-Eur. root dhergh-, ‘to establish’; in-gens, lit. ‘unknown’ (Engl. uncouth), from the root gen-, to know. When used as Nouns the Neuter often appears, e.g. lectum, a bed, tectum, a roof, fatum, destiny, lit. ‘something spoken,’ from the old phrase fari fatum alicui, to lay a doom or spell on one, and (especially in the case of Abstract Nouns) the Feminine, e.g. offensa, repulsa.

[14]: So Engl. friend is properly a Pres. Part. (Goth. frijōnds, ‘loving’).

[15]: Sons has the ONT-stem, a form of the Stem of which traces occur in Athematic Verbs in Latin like eo, volo (euntis Gen., beside iens Nom.; voluntas beside volens).

[16]: Varro tells us the word had this meaning in his native district.

The Past Part. Passive took, as a rule, in I.-Eur. the weak grade (ch. x. § 12) of the Verb-root, e.g. clŭ-to- from cleu-, to hear (Gk. κλυτός, Lat. in-clutus); Lat. dŭc-tus from dūco. The Perfect Ind. Act. has however sometimes influenced its vocalism. The same influence caused the substitution of -sus for -tus, the s-form being originally and properly confined to Dental Verb-stems, e.g. tensus for *tend-tus, ūsus (older ūssus) for *ut-tus, flexus for *flect-tus, salsus for *sald-tus, perculsus for *per-cĕld-tus from the root celd-, to strike (cf. clādes) [dt and tt became by the Latin Law of Sound ss, ch. x. § 17]. But where the Perfect Ind. took s or x, this consonant was given by false analogy to the Part., e.g. farsus from farcio, Perf. farsi for *farc-si; fixus from figo, Perf. fixi; tersus from tergo, Perf. tersi, just as the analogy of haesi produced the late form haesurus and hausi (P.P.P. haustus) hausurus beside hausturus. But in the earlier literature we have the forms with -tus, e.g. tertus. (So in Plautus pulto for class. pulso, § 2.)

The Gerundive (Adj.) in -ndo- has beside it a Gerund (Neut. Noun) in -ndo-, which seems to stand to the Gerundive in the same relation as an Impersonal to a Personal Verb, eundum est in antiquam silvam being Impersonal like itur in antiquam silvam. The Adjectival use seems to have been the original one. [17] In the older Latin writers, when this formation is turned into finite form, i.e. when a statement is made by means of it, the usual method is to employ the Gerund with est governing an object, e.g. agitandum est vigilias, imperandum est servis, carendum est urbe; but in the classical Latin the Gerundive is preferred if the Verb is one which governs the Accusative, e.g. agitandae sunt vigiliae, but still as before, imperandum est servis, carendum est urbe; and the transition from the impersonal to the personal mode of expression, marked by a construction like Plautus’ nominandi istorum copia (a construction allowed by Cicero with a Gen. Pl. for the sake of euphony, e.g. facultas agrorum condonandi) is like the transition from ‘factum itur contumeliam’ to Cato’s ‘contumelia factum itur’ (§ 15 n.).

[17]: Curiously enough the construction of the Gerundive Gen. of Purpose, e.g. Tac. Ann. ii. 59 Germanicus Aegyptum proficiscitur cognoscendae antiquitatis, occurs not only in the early Latin writers, e.g. Lucilius, but also was a usage of the Umbrian language, so that its antiquity cannot be denied.

In the third and fourth Conjugations we find in Early Lat. both -ondo- (e.g. agundus) and -endo- (e.g. agendus), the latter being selected as the classical form (but secundus beside sequendus). This agondo- or agendo- seems to be formed from *agom, Acc. Sg. of a Verbal Noun, and do-, a Verbal Adj. stem meaning ‘giving’ or ‘causing.’ Similarly luendus stands for *luem-dus (cf. lues), curandus stands for *curam-dus (cf. cura), rubendus for *rubem-dus (cf. rubē-facio). The combination of Active and Passive sense which is so peculiar a feature of the Latin Gerundive forms, e.g. ‘agitandum est vigilias’ Act., and ‘agitandae sunt vigiliae’ Pass., may be explained from the double sense that can be attached to an expression like ruborem dare, (1) to blush, Neut., (2) to cause to blush, Act.

With the Gerundive suffix are evidently connected the suffixes of Adjectives in -bundo-, -cundo-, &c., e.g. errabundus, iracundus, facundus (from fari), rubicundus, rotundus. The b of the first of these seems to belong to the root bheu- of fui, &c., the c of the second to the Diminutive suffix seen in rubi-care, albi-care, &c.

The I.-Eur. Person-endings were slightly different in Primary Tenses (the Present Ind., Future Ind., &c.) and in Secondary Tenses (the Preterites Ind., the Tenses of the Optative Mood, &c.). Thus -ti was the 3 Sg. Primary suffix, -t the 3 Sg. Secondary suffix of the Active Voice. In the Perfect Tense an entirely different set of Endings was in use, e.g. -ai 1 Sg. Act., -tha 2 Sg., &c.; and in the Imperative the persons were often distinguished by the addition of Particles (see § 14).

In Passive and Deponent Verbs, Latin departs widely from the I.-Eur. scheme of Passive or Middle Person-endings (contrast Lat. feror, sequor with Gk. φέρο-μαι, ἕπο-μαι, ferimur, sequimur with Gk. φερό-μεθα, ἑπό-μεθα). Latin, as well as the kindred languages of ancient Italy and the Celtic family of languages, uses as the characteristic mark of its passive and deponent flexion the letter r (cf. Old Irish sĕchur 1 Sg., sechĕthar 3 Sg., sechemmar 1 Pl., sechetar 3 Pl. with Lat. sequor, sequitur, sequimur, sequuntur). This r cannot be connected with the Reflexive Pronoun swe- (ch. v. § 1) (Lat. sē Acc.), seeing that s between vowels does not become r in the Celtic languages as in Latin (ch. x. § 19), so that Irish sechur could not represent a form sequo-sē. On the early Impersonal use of these r-forms, see § 6.

1 Sg. The Primary suffix of the Athematic Conjugation (§ 1) was -mī (e.g. Gk. τίθη-μι, εἰμί, I am, for *ἐσ-μί, Lat. sum); in the Thematic the 1 Pers. ended in -ō (e.g. Gk. λέγω, Lat. lego). The suffix in the secondary Tenses of both Conjugations was -m, e.g. Gk. ἔ-φερο-ν, for *ἔφερομ, Lat. eram, sim (an Optative, § 13), ama-bam, amaveram, &c. The Latin Perfect shows the I.-Eur. Middle ending -ai, e.g. dedī, older dedei, vīdī older veidei. (On -ei from earlier -ai, see ch. ii. § 14.)

2 Sg. -si and -s were the I.-Eur. suffixes (e.g. Gk. τίθη-ς, ἔ-φερε-ς, Hom. Gk. ἔσ-σί; Lat. es [earlier *ess, ch. ii. § 9], agis [earlier *ages, ch. ii. § 14], eras). The Latin Perfect shows -stī, older -stei, e.g. dedisti. (On the Imperative 2 Sg. Act., see § 14.)

3 Sg. -ti (Primary) and -t (Secondary) were the I.-Eur. endings (e.g. Gk. ἔσ-τι, τίθη-σι for τίθη-τι, ἔ-φερε for *ἔ-φερε-τ; Lat. est, agit [earlier *aget, ch. ii. § 14], erat). The Secondary ending often appears in Early Latin as d, e.g. feced ‘fecit,’ sied ‘sit’ on the Dvenos inscription (ch. i. § 3 n.). In the Latin Perfect the ending was -eit (class. -ĭt, ch. ii. § 16), being the I.-Eur. Middle ending -ai (Lat. -ei) with t added to distinguish it from the First Person which likewise ended in -ai (Lat. -ei). The long quantity -īt is found in Plautus and the older writers; and even in Ovid we have it after i in the compounds of eo, e.g. interiīt, abiīt, rediīt. (On the Imperative 3 Sg. Act., see § 14.)

1 Plur. In Latin we have in all tenses the ending -mŏs (class. -mus, ch. ii. § 14), while in the other I.-Eur. languages we have a variety of endings, e.g. Att. Gk. φέρο-μεν, ἐ-φέρο-μεν, Dor. Gk. φέρο-μες, ἐ-φέρο-μες.

2 Plur. The ending -tĕ of Gk. φέρε-τε, ἐ-φέρε-τε, &c., appears in Latin only in the Imperative, e.g. fer-te. (On the 2 Plur. of the Future Imperative in -tōtĕ, e.g. fertote, see § 14.) Elsewhere it was replaced by -tĭs (older -tĕs, ch. ii. § 14), e.g. fer-tis (cf. dedistis).

3 Plur. The I.-Eur. suffixes end with -i in Primary but not in Secondary Tenses, viz. -ŏnti (-enti, -ṇti, ch. x. § 12) and -ŏnt (-ent, -ṇt). The form tremonti for tremunt is quoted from an interesting passage, but one of doubtful reading, from the Carmen Saliare:

Cumne (?) tonas, Leucesie, prai tet tremonti,

‘cum tonas, Leucesie, praetremunt te,’ and -ont is often found in the early literature and inscriptions, e.g. cosentiont, ‘consentiunt’ on a Scipio epitaph (ch. iv. § 3 n.). The endings of the Perfect, -ērunt and -ēre (earlier -ērĕ) are difficult to explain. The Imperative seems to add the same particle *tōd, as is seen in the 2 and 3 Sg. (§ 14), to a 3 Pl. form, e.g. ferunto for *feront-tōd.

In Old Latin we find 3 Pl. Pres. Ind. forms in -nunt, e.g. dănunt ‘dant,’ nequīnont ‘nequeunt,’ explēnunt ‘explent,’ which have been explained by the theory that the 3 Pl. of the Pres. Ind. had once ended in -n, *dan, *nequīn, *explēn, and that these forms were a second time provided with a 3 Pl. suffix, much as Greek εἶ, thou art, for *ἐσι, took a fresh 2 Sg. suffix and became εἶς.

1 Sg. The Italo-Celtic ending was -ŏr in the Pres. Ind., e.g. O. Lat. sequōr, class. sequŏr (ch. ii. § 16), O. Ir. sĕchur, apparently an addition of Passive -r (§ 6) to the Active ending -ō (§ 20). Those Tenses and Moods which in Latin formed their 1 Sg. Act. in -m substitute in the Passive -r for -m, e.g. fera-r Fut. and Subj., fereba-r Impft., except in the Perfect group, where a periphrastic form is used, e.g. latus sim, latus essem, not ‘tulerir,’ ‘tulisser’; also latus ero, not ‘tuleror.’

2 Sg. The Latin endings are (1) -rĕ, the same as the Imperative 2 Sg., from I.-Eur. -sŏ (e.g. Gk. ἕπου for ἕπεο from ἕπε-σο, Lat. seque-re, p. 31 n.); that is the usual form in Old Latin and even in Cicero; (2) -rĭs, which adds to this the ending -s of the 2 Sg. Act., *sequerĕ-s becoming sequeris (ch. ii. § 14). Some think -rĕ a weakening of -rĭs. But -ĭs did not become -ĕ in Latin, e.g. militaris. Isolated spellings like TRIBVNOS MILITARE for militaris on an old inscription are not sufficient proofs of this change.

3 Sg. The Italo-Celtic ending is -tŏr (e.g. Lat. sequitur from *sequetŏr, O. Ir. sechĕthar), formed by adding Passive -r to the I.-Eur. Secondary ending -tŏ (e.g. Gk. ἐ-φέρε-το). The Imperative changes to -r the -d of the Particle -tōd, which it appends to its bare stem in the Active (§ 14), e.g. fertor Pass. beside ferto(d) Act.

1 Plur. The Italo-Celtic ending is -mŏr (e.g. Lat. sequimur for *sequŏmŏr, O. Ir. sechemmar), formed by changing to r the s of the Active -mŏs (§ 20).

2 Plur. In Latin the Nom. Plur. of the old Pres. Part. Pass. is used with ellipse of estis, e.g. ferimini (Gk. φερόμενοι, § 17), in the Present Tense and analogical formations in the others, e.g. fereba-mini, fera-mini, ferre-mini. The 2 Pl. Imper. ferimini may be the same with ellipse of este, or an old Infinitive (Gk. φερέμεναι) used in Imperatival sense.

3 Plur. The Italo-Celtic ending is -ntŏr (e.g. Lat. sequuntur from *sequŏntŏr, O. Ir. sechetar), formed by adding Passive r to the I.-Eur. Secondary ending -ntŏ (Gk. ἐ-φέρο-ντο). (Cf. Gk. ἐ-πίμπλη-ντο with Lat. im-ple-ntur). In the Imperative the d of the particle -tōd appended in the 3 Pl. Act. (§ 20) is changed to -r, e.g. feruntor Pass. beside ferunto(d) Act.

Active

I. Indicative

Person	Singular	Plural
1	-m(i) (Athematic Primary), e.g. sum.

-ō (Thematic Primary), e.g. lego.

-m (Secondary), e.g. eram, legebam. | -mŏs, class. -mŭs, e.g. sumus, legimus; eramus, legebamus. | | 2 | -s(i) (Primary), e.g. es for *es-s(i), legis from *legĕs-(i).

-s (Secondary), e.g. eras, legebas. | -tĕs, class. -tĭs, e.g. estis, legitis; eratis, legebatis. | | 3 | -t(i) (Primary), e.g. est, legit from *legĕt(i).

-t (Secondary), e.g. erat, legebat. | -nt(i) (Primary), e.g. sunt, legunt from legŏnti.

-nt (Secondary), e.g. erant, legebant. (But see ch. x. § 17.) |

II. Imperative

Person	Singular	Plural
1	The Subjunctive is used, e.g. legam.	The Subjunctive is used, e.g. legamus.
2 Pres.	The Verb-stem is used, e.g. es (Athem.), lege (Them.).	-tĕ, e.g. es-te; legi-te.
2 Fut.	Adds tō(d) ‘thereupon,’ e.g. es-tō, legitō from *lege-tod.	-tōte, e.g. es-tote, legi-tote.
3	Same as 2 Pers. Fut.	-ntō from -ntō(d), e.g. sunto, legunto.

Passive and Middle

I. Indicative

Person	Singular	Plural
1	-ŏr (Act. -ō), class. -or, e.g. legor.

-r (Act. -m), e.g. legebar. | -mŏr, class. -mŭr, e.g. legimur. | | 2 | (1) -rĕ from -sŏ, e.g. legere.

(2) -rĭs (-rĕ with added -s), e.g. legeris. | Nom. Pl. Masc. of Pres. Part. Pass. is used, e.g. legimini. | | 3 | -tŏr, class. -tŭr, e.g. legitur. | -ntŏr, class. -ntŭr, e.g. leguntur. |

II. Imperative

Person	Singular	Plural
1	The Subjunctive is used, e.g. legar.	The Subjunctive is used, e.g. legamur.
2	-rĕ from -sŏ, e.g. legere.	Nom. Plur. Masc. of the Pres. Part. Pass., e.g. legimini.
3	-tōr, class. -tŏr, e.g. legitor.	-ntōr, class. -ntŏr, e.g. leguntor.

The irregularity of many verbs consists in their use of different roots for different tenses. The Substantive Verb shows the root ES- in some tenses, the root BHEU- in others, e.g. Perf. fui; and the heteroclite conjugation of verbs like ‘to be,’ ‘to go,’ seems to date from the I.-Eur. period. Other Latin examples are fio (for *fuvio, from bhw-, a weakened form of the root bheu-) and factus sum; ferio and percussi; fero and tuli (from the root of tollo, Gk. τετλάναι); tollo (for *tol-no) and sustuli (Compound with Prep. sub, subs); vescor and pastus sum; arguor and convictus sum; reminiscor and recordatus sum; medeor and medicatus sum; surgo differs from surrexi and surrectus in being syncopated (cf. porgo and porrigo); vis beside volo comes from a root vei-, which had the same meaning as the root vel-, to wish. (On vel-, e.g. velim, and vol-, e.g. volo, see ch. ii. § 8.)

Other verbs, classed as Irregular, are the Defective Verbs: coepi, the Present of which, coepio, is found in O. Lat., e.g. neque ego lites coepio, Plaut., and is a compound of cum and apio, to fasten, whence apiscor and aptus; inquam, perhaps a Subj.-Fut., ‘I will say,’ from the root seqᵘ-, to speak (cf. O. Lat. insēque or insece [18]), other parts of which are Pres. Ind. inquis, inquit, inquiunt and Imper. inque (Plaut.); aio, for *ahio from the root agh-, ‘to say,’ and infit are other Defective Verbs of kindred meaning to inquam. (Cf. our Defect. Vb. ‘quoth.’)

[^18]: Ennius’ translation of Ἄνδρα μοι ἔννεπε, Μοῦσα was: inseque (insece), Musa, uirum.

The irregularity of verbs like sum, eo, volo, &c., consists, as we have seen (§ 1) in their having belonged to the Athematic Conjugation (hence 1 Pl. sumus, volumus with ŭ not ŏ), of which there are not so many remains in Latin as in Greek (e.g. all the Greek verbs in -μι are athematic). Edo, to eat, is thematic in 1 Sg. Pres. Ind. ĕd-o, but athematic in the other persons, which often show a by-form ēd-, e.g. essem Impft. Subj. for *ed-sem, esse Inf. for *ed-se. Fero has been attracted into the Athematic Conj., fer-s 2 Sg., fer-t 3 Sg., ferrem Impft. Subj. for *fer-sem, ferre Inf. for *fer-se, fer Imper. Dō (root DŌ-, in weak form Lat. dă-) was originally athematic and must have been declined *dōm(i), *dō-s, *dō-t, *dă-mos, *dă-tes, dant (cf. Gk. δίδωμι, δίδομεν, and cf. ch. x. § 12); but the ā of dāmus, dātis, &c., has caused its partial transference to the First Conjugation type. (On cĕ-dō, Imperat., see § 14.) It was confused with another -do, meaning ‘to place’ or ‘put’ (root DHĒ-, Gk. τί-θη-μι), to which belong the Compounds condo, ‘to put together,’ subdo, ‘to put under,’ &c., and which must have been declined *con-dē-m(i), *con-dē-s, *con-dē-t, *con-dă-mos, *condă-tes, &c. Here the ĭ to which ă was reduced in the unaccented syllable (ch. ii. § 14), con-di-mus, con-di-tis, &c., caused the transference of condo, subdo, &c., to the Third Conjugation type. I.-Eur. dh when initial became f in Latin (ch. x. § 17), so that the simple verb would be *fē-m(i), *fē-s, with Plur. *fă-mos, &c. A derivative from this is făc-io, with the c of which we may compare the κ of Gk. ἔ-θηκ-α. Sto was another Athematic Verb (root STĀ-, Gk. ἵστημι for *σι-στά-μι, Pl. ἵ-στά-μεν), with an original declension, *stā-m(i), stā-s, stā-t, *stā-mos, *stā-tes, stant, but in a verb of this kind the transference to the First Conjugation was easy. Sto had in Early Latin the transitive sense of Gk. ἵστημι, e.g. med Mano statod, ‘set me (as an offering) to Manus’ (the Good Deity; cf. Mānes, im-mānis, lit. ‘not good’) on the Dvenos inscr. (ch. i. § 1 n.), but in class. Lat. this sense was given to the transitive form si-sto (Gk. ἵστημι for *σι-στά-μι), which appears as a Third Conj. Verb.

Compounds of these Athematic Verbs are possum, nōlo, mālo. Possum is found in the older writers in its uncompounded form potis sum and pote sum, e.g. potis est Lucr. i. 452, potesse i. 665. Pote is properly the Neuter of potis, e.g. pote est (class. potest), ‘it is possible,’ but the Masc. (and Fem.) and Neut. forms are used of any gender and of any number, e.g. potis est, ‘it is possible’ Ter. Phorm. 379; credo equidem potis esse te, scelus Plaut. Possum stands for pote-sum (on the Syncope of -ĕ see ch. ii. § 12), possim for pote-sim, &c.; and possem, posse (for potessem, potesse) have followed their analogy. Mālo, a contraction of māvolo (Plaut.), comes similarly from măge (a by-form of magis, which was wrongly treated as an Adj. magis with Neut. mage) united into a word-group (ch. ii. § 11) with vŏlo; and nōlo from the Negative nĕ- and vŏlo. *Măg(e)-volo became mavvolo, written māvolo; *nĕ-volo (O. Lat. nē-vis, nē-volt) became *nāvolo (ch. x. § 4) and with loss of v between vowels (ch. ii. § 12) nōlo. The Imperat. nōlī may be due to the use of nōlīs (Opt.) in Imperatival sense, the -s being dropped to suit the type of the Imperative 2 Sg.

§21

(β) Passive (Deponent).

(β) Passive (Deponent).

1 Sg. The Italo-Celtic ending was -ŏr in the Pres. Ind., e.g. O. Lat. sequōr, class. sequŏr (ch. ii. § 16), O. Ir. sĕchur, apparently an addition of Passive -r (§ 6) to the Active ending -ō (§ 20). Those Tenses and Moods which in Latin formed their 1 Sg. Act. in -m substitute in the Passive -r for -m, e.g. fera-r Fut. and Subj., fereba-r Impft., except in the Perfect group, where a periphrastic form is used, e.g. latus sim, latus essem, not ‘tulerir,’ ‘tulisser’; also latus ero, not ‘tuleror.’

2 Sg. The Latin endings are (1) -rĕ, the same as the Imperative 2 Sg., from I.-Eur. -sŏ (e.g. Gk. ἕπου for ἕπεο from ἕπε-σο, Lat. seque-re, p. 31 n.); that is the usual form in Old Latin and even in Cicero; (2) -rĭs, which adds to this the ending -s of the 2 Sg. Act., *sequerĕ-s becoming sequeris (ch. ii. § 14). Some think -rĕ a weakening of -rĭs. But -ĭs did not become -ĕ in Latin, e.g. militaris. Isolated spellings like TRIBVNOS MILITARE for militaris on an old inscription are not sufficient proofs of this change.

3 Sg. The Italo-Celtic ending is -tŏr (e.g. Lat. sequitur from *sequetŏr, O. Ir. sechĕthar), formed by adding Passive -r to the I.-Eur. Secondary ending -tŏ (e.g. Gk. ἐ-φέρε-το). The Imperative changes to -r the -d of the Particle -tōd, which it appends to its bare stem in the Active (§ 14), e.g. fertor Pass. beside ferto(d) Act.

1 Plur. The Italo-Celtic ending is -mŏr (e.g. Lat. sequimur for *sequŏmŏr, O. Ir. sechemmar), formed by changing to r the s of the Active -mŏs (§ 20).

2 Plur. In Latin the Nom. Plur. of the old Pres. Part. Pass. is used with ellipse of estis, e.g. ferimini (Gk. φερόμενοι, § 17), in the Present Tense and analogical formations in the others, e.g. fereba-mini, fera-mini, ferre-mini. The 2 Pl. Imper. ferimini may be the same with ellipse of este, or an old Infinitive (Gk. φερέμεναι) used in Imperatival sense.

3 Plur. The Italo-Celtic ending is -ntŏr (e.g. Lat. sequuntur from *sequŏntŏr, O. Ir. sechetar), formed by adding Passive r to the I.-Eur. Secondary ending -ntŏ (Gk. ἐ-φέρο-ντο). (Cf. Gk. ἐ-πίμπλη-ντο with Lat. im-ple-ntur). In the Imperative the d of the particle -tōd appended in the 3 Pl. Act. (§ 20) is changed to -r, e.g. feruntor Pass. beside ferunto(d) Act.

1 Sg. The Italo-Celtic ending was -ŏr in the Pres. Ind., e.g. O. Lat. sequōr, class. sequŏr (ch. ii. § 16), O. Ir. sĕchur, apparently an addition of Passive -r (§ 6) to the Active ending -ō (§ 20). Those Tenses and Moods which in Latin formed their 1 Sg. Act. in -m substitute in the Passive -r for -m, e.g. fera-r Fut. and Subj., fereba-r Impft., except in the Perfect group, where a periphrastic form is used, e.g. latus sim, latus essem, not ‘tulerir,’ ‘tulisser’; also latus ero, not ‘tuleror.’

2 Sg. The Latin endings are (1) -rĕ, the same as the Imperative 2 Sg., from I.-Eur. -sŏ (e.g. Gk. ἕπου for ἕπεο from ἕπε-σο, Lat. seque-re, p. 31 n.); that is the usual form in Old Latin and even in Cicero; (2) -rĭs, which adds to this the ending -s of the 2 Sg. Act., *sequerĕ-s becoming sequeris (ch. ii. § 14). Some think -rĕ a weakening of -rĭs. But -ĭs did not become -ĕ in Latin, e.g. militaris. Isolated spellings like TRIBVNOS MILITARE for militaris on an old inscription are not sufficient proofs of this change.

3 Sg. The Italo-Celtic ending is -tŏr (e.g. Lat. sequitur from *sequetŏr, O. Ir. sechĕthar), formed by adding Passive -r to the I.-Eur. Secondary ending -tŏ (e.g. Gk. ἐ-φέρε-το). The Imperative changes to -r the -d of the Particle -tōd, which it appends to its bare stem in the Active (§ 14), e.g. fertor Pass. beside ferto(d) Act.

1 Plur. The Italo-Celtic ending is -mŏr (e.g. Lat. sequimur for *sequŏmŏr, O. Ir. sechemmar), formed by changing to r the s of the Active -mŏs (§ 20).

2 Plur. In Latin the Nom. Plur. of the old Pres. Part. Pass. is used with ellipse of estis, e.g. ferimini (Gk. φερόμενοι, § 17), in the Present Tense and analogical formations in the others, e.g. fereba-mini, fera-mini, ferre-mini. The 2 Pl. Imper. ferimini may be the same with ellipse of este, or an old Infinitive (Gk. φερέμεναι) used in Imperatival sense.

3 Plur. The Italo-Celtic ending is -ntŏr (e.g. Lat. sequuntur from *sequŏntŏr, O. Ir. sechetar), formed by adding Passive r to the I.-Eur. Secondary ending -ntŏ (Gk. ἐ-φέρο-ντο). (Cf. Gk. ἐ-πίμπλη-ντο with Lat. im-ple-ntur). In the Imperative the d of the particle -tōd appended in the 3 Pl. Act. (§ 20) is changed to -r, e.g. feruntor Pass. beside ferunto(d) Act.

§22

Scheme of the Latin Person-endings.

Scheme of the Latin Person-endings.

Active

I. Indicative

Person	Singular	Plural
1	-m(i) (Athematic Primary), e.g. sum.

-ō (Thematic Primary), e.g. lego.

-m (Secondary), e.g. eram, legebam. | -mŏs, class. -mŭs, e.g. sumus, legimus; eramus, legebamus. | | 2 | -s(i) (Primary), e.g. es for *es-s(i), legis from *legĕs-(i).

-s (Secondary), e.g. eras, legebas. | -tĕs, class. -tĭs, e.g. estis, legitis; eratis, legebatis. | | 3 | -t(i) (Primary), e.g. est, legit from *legĕt(i).

-t (Secondary), e.g. erat, legebat. | -nt(i) (Primary), e.g. sunt, legunt from legŏnti.

-nt (Secondary), e.g. erant, legebant. (But see ch. x. § 17.) |

II. Imperative

Person	Singular	Plural
1	The Subjunctive is used, e.g. legam.	The Subjunctive is used, e.g. legamus.
2 Pres.	The Verb-stem is used, e.g. es (Athem.), lege (Them.).	-tĕ, e.g. es-te; legi-te.
2 Fut.	Adds tō(d) ‘thereupon,’ e.g. es-tō, legitō from *lege-tod.	-tōte, e.g. es-tote, legi-tote.
3	Same as 2 Pers. Fut.	-ntō from -ntō(d), e.g. sunto, legunto.

Passive and Middle

I. Indicative

Person	Singular	Plural
1	-ŏr (Act. -ō), class. -or, e.g. legor.

-r (Act. -m), e.g. legebar. | -mŏr, class. -mŭr, e.g. legimur. | | 2 | (1) -rĕ from -sŏ, e.g. legere.

(2) -rĭs (-rĕ with added -s), e.g. legeris. | Nom. Pl. Masc. of Pres. Part. Pass. is used, e.g. legimini. | | 3 | -tŏr, class. -tŭr, e.g. legitur. | -ntŏr, class. -ntŭr, e.g. leguntur. |

II. Imperative

Person	Singular	Plural
1	The Subjunctive is used, e.g. legar.	The Subjunctive is used, e.g. legamur.
2	-rĕ from -sŏ, e.g. legere.	Nom. Plur. Masc. of the Pres. Part. Pass., e.g. legimini.
3	-tōr, class. -tŏr, e.g. legitor.	-ntōr, class. -ntŏr, e.g. leguntor.

Active

I. Indicative

Person	Singular	Plural
1	-m(i) (Athematic Primary), e.g. sum. -ō (Thematic Primary), e.g. lego. -m (Secondary), e.g. eram, legebam.	-mŏs, class. -mŭs, e.g. sumus, legimus; eramus, legebamus.
2	-s(i) (Primary), e.g. es for es-s(i), legis from legĕs-(i). -s (Secondary), e.g. eras, legebas.	-tĕs, class. -tĭs, e.g. estis, legitis; eratis, legebatis.
3	-t(i) (Primary), e.g. est, legit from legĕt(i). -t (Secondary), e.g. erat, legebat*.	-nt(i) (Primary), e.g. sunt, legunt from legŏnti. -nt (Secondary), e.g. erant, legebant. (But see ch. x. § 17.)

II. Imperative

Person	Singular	Plural
1	The Subjunctive is used, e.g. legam.	The Subjunctive is used, e.g. legamus.
2 Pres.	The Verb-stem is used, e.g. es (Athem.), lege (Them.).	-tĕ, e.g. es-te; legi-te.
2 Fut.	Adds tō(d) ‘thereupon,’ e.g. es-tō, legitō from *lege-tod.	-tōte, e.g. es-tote, legi-tote.
3	Same as 2 Pers. Fut.	-ntō from -ntō(d), e.g. sunto, legunto.

Passive and Middle

I. Indicative

Person	Singular	Plural
1	-ŏr (Act. -ō), class. -or, e.g. legor. -r (Act. -m), e.g. legebar.	-mŏr, class. -mŭr, e.g. legimur.
2	(1) -rĕ from -sŏ, e.g. legere. (2) -rĭs (-rĕ with added -s), e.g. legeris.	Nom. Pl. Masc. of Pres. Part. Pass. is used, e.g. legimini.
3	-tŏr, class. -tŭr, e.g. legitur.	-ntŏr, class. -ntŭr, e.g. leguntur.

II. Imperative

Person	Singular	Plural
1	The Subjunctive is used, e.g. legar.	The Subjunctive is used, e.g. legamur.
2	-rĕ from -sŏ, e.g. legere.	Nom. Plur. Masc. of the Pres. Part. Pass., e.g. legimini.
3	-tōr, class. -tŏr, e.g. legitor.	-ntōr, class. -ntŏr, e.g. leguntor.

§23

Some Irregular Verbs.

Some Irregular Verbs.

The irregularity of many verbs consists in their use of different roots for different tenses. The Substantive Verb shows the root ES- in some tenses, the root BHEU- in others, e.g. Perf. fui; and the heteroclite conjugation of verbs like ‘to be,’ ‘to go,’ seems to date from the I.-Eur. period. Other Latin examples are fio (for *fuvio, from bhw-, a weakened form of the root bheu-) and factus sum; ferio and percussi; fero and tuli (from the root of tollo, Gk. τετλάναι); tollo (for *tol-no) and sustuli (Compound with Prep. sub, subs); vescor and pastus sum; arguor and convictus sum; reminiscor and recordatus sum; medeor and medicatus sum; surgo differs from surrexi and surrectus in being syncopated (cf. porgo and porrigo); vis beside volo comes from a root vei-, which had the same meaning as the root vel-, to wish. (On vel-, e.g. velim, and vol-, e.g. volo, see ch. ii. § 8.)

Other verbs, classed as Irregular, are the Defective Verbs: coepi, the Present of which, coepio, is found in O. Lat., e.g. neque ego lites coepio, Plaut., and is a compound of cum and apio, to fasten, whence apiscor and aptus; inquam, perhaps a Subj.-Fut., ‘I will say,’ from the root seqᵘ-, to speak (cf. O. Lat. insēque or insece [18]), other parts of which are Pres. Ind. inquis, inquit, inquiunt and Imper. inque (Plaut.); aio, for *ahio from the root agh-, ‘to say,’ and infit are other Defective Verbs of kindred meaning to inquam. (Cf. our Defect. Vb. ‘quoth.’)

[^18]: Ennius’ translation of Ἄνδρα μοι ἔννεπε, Μοῦσα was: inseque (insece), Musa, uirum.

The irregularity of verbs like sum, eo, volo, &c., consists, as we have seen (§ 1) in their having belonged to the Athematic Conjugation (hence 1 Pl. sumus, volumus with ŭ not ŏ), of which there are not so many remains in Latin as in Greek (e.g. all the Greek verbs in -μι are athematic). Edo, to eat, is thematic in 1 Sg. Pres. Ind. ĕd-o, but athematic in the other persons, which often show a by-form ēd-, e.g. essem Impft. Subj. for *ed-sem, esse Inf. for *ed-se. Fero has been attracted into the Athematic Conj., fer-s 2 Sg., fer-t 3 Sg., ferrem Impft. Subj. for *fer-sem, ferre Inf. for *fer-se, fer Imper. Dō (root DŌ-, in weak form Lat. dă-) was originally athematic and must have been declined *dōm(i), *dō-s, *dō-t, *dă-mos, *dă-tes, dant (cf. Gk. δίδωμι, δίδομεν, and cf. ch. x. § 12); but the ā of dāmus, dātis, &c., has caused its partial transference to the First Conjugation type. (On cĕ-dō, Imperat., see § 14.) It was confused with another -do, meaning ‘to place’ or ‘put’ (root DHĒ-, Gk. τί-θη-μι), to which belong the Compounds condo, ‘to put together,’ subdo, ‘to put under,’ &c., and which must have been declined *con-dē-m(i), *con-dē-s, *con-dē-t, *con-dă-mos, *condă-tes, &c. Here the ĭ to which ă was reduced in the unaccented syllable (ch. ii. § 14), con-di-mus, con-di-tis, &c., caused the transference of condo, subdo, &c., to the Third Conjugation type. I.-Eur. dh when initial became f in Latin (ch. x. § 17), so that the simple verb would be *fē-m(i), *fē-s, with Plur. *fă-mos, &c. A derivative from this is făc-io, with the c of which we may compare the κ of Gk. ἔ-θηκ-α. Sto was another Athematic Verb (root STĀ-, Gk. ἵστημι for *σι-στά-μι, Pl. ἵ-στά-μεν), with an original declension, *stā-m(i), stā-s, stā-t, *stā-mos, *stā-tes, stant, but in a verb of this kind the transference to the First Conjugation was easy. Sto had in Early Latin the transitive sense of Gk. ἵστημι, e.g. med Mano statod, ‘set me (as an offering) to Manus’ (the Good Deity; cf. Mānes, im-mānis, lit. ‘not good’) on the Dvenos inscr. (ch. i. § 1 n.), but in class. Lat. this sense was given to the transitive form si-sto (Gk. ἵστημι for *σι-στά-μι), which appears as a Third Conj. Verb.

Compounds of these Athematic Verbs are possum, nōlo, mālo. Possum is found in the older writers in its uncompounded form potis sum and pote sum, e.g. potis est Lucr. i. 452, potesse i. 665. Pote is properly the Neuter of potis, e.g. pote est (class. potest), ‘it is possible,’ but the Masc. (and Fem.) and Neut. forms are used of any gender and of any number, e.g. potis est, ‘it is possible’ Ter. Phorm. 379; credo equidem potis esse te, scelus Plaut. Possum stands for pote-sum (on the Syncope of -ĕ see ch. ii. § 12), possim for pote-sim, &c.; and possem, posse (for potessem, potesse) have followed their analogy. Mālo, a contraction of māvolo (Plaut.), comes similarly from măge (a by-form of magis, which was wrongly treated as an Adj. magis with Neut. mage) united into a word-group (ch. ii. § 11) with vŏlo; and nōlo from the Negative nĕ- and vŏlo. *Măg(e)-volo became mavvolo, written māvolo; *nĕ-volo (O. Lat. nē-vis, nē-volt) became *nāvolo (ch. x. § 4) and with loss of v between vowels (ch. ii. § 12) nōlo. The Imperat. nōlī may be due to the use of nōlīs (Opt.) in Imperatival sense, the -s being dropped to suit the type of the Imperative 2 Sg.

The irregularity of many verbs consists in their use of different roots for different tenses. The Substantive Verb shows the root ES- in some tenses, the root BHEU- in others, e.g. Perf. fui; and the heteroclite conjugation of verbs like ‘to be,’ ‘to go,’ seems to date from the I.-Eur. period. Other Latin examples are fio (for *fuvio, from bhw-, a weakened form of the root bheu-) and factus sum; ferio and percussi; fero and tuli (from the root of tollo, Gk. τετλάναι); tollo (for *tol-no) and sustuli (Compound with Prep. sub, subs); vescor and pastus sum; arguor and convictus sum; reminiscor and recordatus sum; medeor and medicatus sum; surgo differs from surrexi and surrectus in being syncopated (cf. porgo and porrigo); vis beside volo comes from a root vei-, which had the same meaning as the root vel-, to wish. (On vel-, e.g. velim, and vol-, e.g. volo, see ch. ii. § 8.)

Other verbs, classed as Irregular, are the Defective Verbs: coepi, the Present of which, coepio, is found in O. Lat., e.g. neque ego lites coepio, Plaut., and is a compound of cum and apio, to fasten, whence apiscor and aptus; inquam, perhaps a Subj.-Fut., ‘I will say,’ from the root seqᵘ-, to speak (cf. O. Lat. insēque or insece [18]), other parts of which are Pres. Ind. inquis, inquit, inquiunt and Imper. inque (Plaut.); aio, for *ahio from the root agh-, ‘to say,’ and infit are other Defective Verbs of kindred meaning to inquam. (Cf. our Defect. Vb. ‘quoth.’)

[^18]: Ennius’ translation of Ἄνδρα μοι ἔννεπε, Μοῦσα was: inseque (insece), Musa, uirum.

The irregularity of verbs like sum, eo, volo, &c., consists, as we have seen (§ 1) in their having belonged to the Athematic Conjugation (hence 1 Pl. sumus, volumus with ŭ not ŏ), of which there are not so many remains in Latin as in Greek (e.g. all the Greek verbs in -μι are athematic). Edo, to eat, is thematic in 1 Sg. Pres. Ind. ĕd-o, but athematic in the other persons, which often show a by-form ēd-, e.g. essem Impft. Subj. for *ed-sem, esse Inf. for *ed-se. Fero has been attracted into the Athematic Conj., fer-s 2 Sg., fer-t 3 Sg., ferrem Impft. Subj. for *fer-sem, ferre Inf. for *fer-se, fer Imper. Dō (root DŌ-, in weak form Lat. dă-) was originally athematic and must have been declined *dōm(i), *dō-s, *dō-t, *dă-mos, *dă-tes, dant (cf. Gk. δίδωμι, δίδομεν, and cf. ch. x. § 12); but the ā of dāmus, dātis, &c., has caused its partial transference to the First Conjugation type. (On cĕ-dō, Imperat., see § 14.) It was confused with another -do, meaning ‘to place’ or ‘put’ (root DHĒ-, Gk. τί-θη-μι), to which belong the Compounds condo, ‘to put together,’ subdo, ‘to put under,’ &c., and which must have been declined *con-dē-m(i), *con-dē-s, *con-dē-t, *con-dă-mos, *condă-tes, &c. Here the ĭ to which ă was reduced in the unaccented syllable (ch. ii. § 14), con-di-mus, con-di-tis, &c., caused the transference of condo, subdo, &c., to the Third Conjugation type. I.-Eur. dh when initial became f in Latin (ch. x. § 17), so that the simple verb would be *fē-m(i), *fē-s, with Plur. *fă-mos, &c. A derivative from this is făc-io, with the c of which we may compare the κ of Gk. ἔ-θηκ-α. Sto was another Athematic Verb (root STĀ-, Gk. ἵστημι for *σι-στά-μι, Pl. ἵ-στά-μεν), with an original declension, *stā-m(i), stā-s, stā-t, *stā-mos, *stā-tes, stant, but in a verb of this kind the transference to the First Conjugation was easy. Sto had in Early Latin the transitive sense of Gk. ἵστημι, e.g. med Mano statod, ‘set me (as an offering) to Manus’ (the Good Deity; cf. Mānes, im-mānis, lit. ‘not good’) on the Dvenos inscr. (ch. i. § 1 n.), but in class. Lat. this sense was given to the transitive form si-sto (Gk. ἵστημι for *σι-στά-μι), which appears as a Third Conj. Verb.

Compounds of these Athematic Verbs are possum, nōlo, mālo. Possum is found in the older writers in its uncompounded form potis sum and pote sum, e.g. potis est Lucr. i. 452, potesse i. 665. Pote is properly the Neuter of potis, e.g. pote est (class. potest), ‘it is possible,’ but the Masc. (and Fem.) and Neut. forms are used of any gender and of any number, e.g. potis est, ‘it is possible’ Ter. Phorm. 379; credo equidem potis esse te, scelus Plaut. Possum stands for pote-sum (on the Syncope of -ĕ see ch. ii. § 12), possim for pote-sim, &c.; and possem, posse (for potessem, potesse) have followed their analogy. Mālo, a contraction of māvolo (Plaut.), comes similarly from măge (a by-form of magis, which was wrongly treated as an Adj. magis with Neut. mage) united into a word-group (ch. ii. § 11) with vŏlo; and nōlo from the Negative nĕ- and vŏlo. *Măg(e)-volo became mavvolo, written māvolo; *nĕ-volo (O. Lat. nē-vis, nē-volt) became *nāvolo (ch. x. § 4) and with loss of v between vowels (ch. ii. § 12) nōlo. The Imperat. nōlī may be due to the use of nōlīs (Opt.) in Imperatival sense, the -s being dropped to suit the type of the Imperative 2 Sg.