i UC-NRLF B 4 03b 238 ■J '^IX^^/S^^W- University of California. FROM THE LIBRARY OF Dr. martin KELLOGG. GIFT OF MRS. LOUISE B. KELLOGG. No. ANALYSIS OF THE LATIN YERB, ILLUSTRATED BY THE FORMS OF THE SANSKRIT. BY CHARLES H. PARKHUEST. BOSTON: GINN BROTHERS AND COIVIPANY. 1870. r Entered, according to Act of Congress, in the year 1S70, by CHARLES H. PARKHURST, In the Clerk's Office of the District Court of the District of Massachusetts. CAMBRIDGE: PRESS OF JOHX WILSO:^ AXD SON. PREFACE. The immediate aim of this treatise is to familiarize the student with the earlier and later forms of the Latin verb, and the method by which the latter have been corrupted from the former. It is not in the power of the Latin, however, to ex^DOund itself. Latin scholarship may state the facts, but it cannot give the philosophy, of Latin formation. The super- ficial analysis of the verb amo, for example, in the first per- son plural of the present subjunctive active would be, — root, am ; connecting vowel, e ; personal ending, mus. An ac- quaintance with Latin, however thorough, will not, without assistance from the cognate languages, discover in that connec- tive e a union of a with i, the former a corruption of aja, the affix of conjugation, and the latter, like the iota of the Greek, the modal sign of the optative. We have preferred, therefore, to explain the Latin system of conjugation by reference to the Sanskrit, not at all because the Sanskrit is the progenitor of the Latin, but because it most fully retains the forms belonging to the parent language of the entire Indo- European family, and as such offers to us the proximate orig- inal, from which the later Latin derivatives are corrupted. The mediate aim of these pages is to introduce the pupil to the study of Comparative Grammar, — a science which is engaging the energies of the profoundest scholars of Germany, and one which, though failing to satisfy the utilitarian ten- dencies of the age, is yet rich in promise to the less ambitious few, with whom research and discovery are their own suffi- cient and abundant reward. IV PREFACE. We have not assumed on the part of the student any acquaintance with either the Sanskrit or the Greek. As a necessary consequence, however, of the sisterly relation existin"; between the Greek and Latin, one familiar with both cannot, we believe, fail to discover in our discussion of the Latin the solution of many problems in Greek formation. In the preparation of this Manual we have taken as our guide the writings of Professors Bopp and Schleicher. As respects the content of the " Analysis " we lay no claim to originality, its primary object being to give the best approved results of German research, in a form convenient and intel- ligible to the English reader. Many of the conclusions attained rest rather upon probable than upon demonstrative evidence. Care has been taken in each instance to put these conclusions in no more positive form than that with which they are enunciated by the authorities we follow. We are happy to acknowledge our indebtedness to Profes- sor Greenough, of Harvard College, for his examination of the manuscript, and for valuable suggestions and criticisms. Conscious of the imperfection of the work both in form and content, we submit it to the public with the hope that it may not be altogether without its influence in quickening the spirit of linguistic inquiry, and that its perusal may serve to reproduce in the case of some earnest student, some- thing of the pleasure experienced by the author in the course of its preparation. WiLLiSTON Seminary, Feb. 24, 1870. ANALYSIS OF THE VERB. 1. The oldest form of a word is called its Root. 2. Roots are of two kinds, Pronominal and Verbal. From pro- nominal roots are developed pronouns and particles ; from verbal roots, verbs and nouns (substantive and adjective). 3. Inflection is the union of pronominal with verbal roots. When so combined as to denote relations of gender, number, and case, the process is termed Declension. When so combined as to denote relations of voice, mood, tense, number, and person, the process is called Conjugation. 4. Etymological analysis is the converse of inflection, and con- sists in resolving a verb or noun into its ultimate verbal and pro- nominal elements. 5 The analysis of the verb consists in separating frohi the root those pronominal accretions which mark distinctions in conjuga- tion, and serve to denote relations of voice, mood, tense, number, and person. The Latin Alphabet. Note. — As preliminary to the analysis proper, sections 6-49 will be devoted to a consideration of some of the more general euphonic and emphatic changes employed in verbal formation. (a.) Vowels. 6. The relative value of the vowels will be best determined by giving to them the sounds Avith which they are pronounced in the continental languages of Europe: viz., a, as m father, man; e, as in they, met; i, as in machine, htm; o, as in no, not; u, as in rule, full. 7. By comparing these sounds, it will appear that only that of a, i, and u is simple, — i.e., produced by the vocal organs in a single 10 ANALYSIS OF THE VERB. fixed position; -while that of e and o Is compound, — i.e., in their production the position of the organs is changed. 8. The first element of o is a; the second, u: hence we place a -{- u = o. The first element of e is a ; the second, i : hence we place a + i=^e- An instance wherein e stands in place of an earlier a + i, will be found, for example, in the present subjunctive of conjugation first ; amem for an earlier amaim ; amemus for an earlier aniaiinus. 9. u and its corresponding semi-vowel v are readily exchange- able, as also i and its semi-vowel j. Thus In the perfect of con- jugation second we shall find mon-(e)-vi becoming mon-ui; and in the affix of conjugation second, aj changed successively to ai and e. 10. As In the Sanskrit, so also In the Latin, uv sometimes develops Itself out of u ; thus fluo becomes In the perfect fluvsi, whence fluxi. 11. The weight of a vowel Is the fulness of tone with which it is enunciated. The order of vowels from heaviest to lightest, ranked according to their weight, is as follows : a, u, o, e, i. The Latin everywhere exhibits a tendency to pass from a heavier to a lighter. 12. As regards the verb, this tendency Is particularly noticeable in the forfnation of reduplicated perfects. Thus fallo, perfect fe-felli for fa-falli; cano, perfect ce-cini for ca-cani; cado, perfect ce-cidi for ca-cadi. Cecini and cecidi also Illustrate another tendency In the formation of reduplicates, that of attenuat- ing the perfect more In Its radical than in Its reduplicate syllable. Thus in place of ca-cani not ce-ceui, but ce-cini ; so ce-cidi instead of ce-cedi, te-tigi for te-tegi. 13. A radical u or o, however, reappears in the perfect, both in its radical and reduplicate syllable. Hence tundo, tu-tudi ;' posco, po-posci. 14. In like manner verbs in composition often lighten the vowel of the root; e.g., concino for con-cano, assideo for as-sedeo. 15. The tendency of a nasal Is to convert the preceding vowel into u; e.g., capiunt for an earlier capiant. 16. The Influence of r, or of any two consonants, is often to change the preceding i Into e; e.g., amaverunt of the perfect indicative for amavirunt, acceptum for acciptum. ANALYSIS OF THE VERB. 11 17. A final m or t tends to shorten the preceding vowel ; e.g., si-m becomes sun; audia-m, audiam; ama-t, amat; rega-t, regat. 18. e at the end of a word is usually prefen-ed to i; e.g., in the imperative of conjugation tliu-d, lege for the regular legi. (6.) Consonants. 19. The consonants of the Latm are nineteen ; seventeen sim- ple, and tAvo, x (=cs) and z (=ds), compound. Consonants produced with a resonance of the vocal cords are termed sonant ; those produced without such resonance, surd. 20. As still farther classified according to the method of their generation, the scheme of the Latin consonants is as follows : — Palatals. Labials. Linguals. - w / Surds. c, k, q, h. p, f. t, s. ^^^ -^ C g, j. b, V. d. SON^^NTS. -j Nasals, {m, ^ n. ^ j^,^^^^^^ 21. As a result of inflection, letters sometimes become so com- bined as to be difficult of pronunciation. The principles accord- in o- to which such combinations are simplified are caUed Laws of Euphonv. The more general of these which require to be applied in the analysis of the verb are the following : — 22. Before the affix si of the perfect a sonant palatal becomes surd: e.g., aug (augeo) + si = auc + si ; reg (rego)+si = rec -f- si. 23. c, or any other surd palatal converted to c, combines with s to form X, according to 19 : e.g., auc + si = auzi ; coq + si = coc -f si = coxi ; veh + si = vec + si =: vexi. 24. Exception. Roots ending in a palatal immediately preceded by 1 or r drop the palatal before the affix si: e.g., algeo, alsi; mergo, mersi. 25. Before si the labial sonant b is represented by its cog- nate (vid. 41, 6) surd p: e.g., nubo, nupsi; scribo, scripsi. 26. V, through an intermediate conversion to c, combines with s to form X; after the analogy of palatals : e.g., vivo, viv-f si = Vic + si = vizi ; fluo, fluv + si (vid. 10) = flue + si = fluxi. 27. Roots in m optionally insert p before the affix si: e.g., como, com-p-si or comsi; sumo, sum-p-si or sumsi. 12 ANALYSIS OF THE VERB. 28. Lingual roots in d, t, and r, cannot dii-ectly coalesce -with the affix s. Such succession of consonants may be prevented (a) by dropi^ing the characteristic: e.g., ardsi from ardeo becomes arsi; sentsi from sentio^ sensi; haersi from haereo, hsesi; (b) by dropping the s: e.g., prandsi from prando becomes prandi ; vertsi from verto, verti ; (c) by assimilating (vid. 41, c) the characteristic to the affix: e.g., cedo, cedsi, cessi; quatio, quatsi, quassi; gero, gersi, gessi. 29. Before the affix t, the sonant palatal g, and the surd palatals q and h, are changed to c: e.g., jungo, jungtum, junctum; coquo, coqtum, coctum; veho, vehtum, vectum. 30. Exception. Roots ending in a palatal immediately preceded by 1 or r commonly affix t (frequently converted into s, its asso- ciate lingual surd) with the elision of the palatal: e.g., fulc (fulcio) becomes in the supine fultum; sarc (sarcio) becomes sartum; mulceo gives mulsum; tergeo, tersum, 31. Before t the labial sonant b is represented by its cognate surd p: e.g., glubo, glubtum, gluptum; nubo, nubtum, nuptum. 32. The harshness of the combination vt may be relieved (a) by the conversion of v into u: e.g., cautum for cavtum, from caveo ; volutum for volvtum, from volvo ; (6) by the elision of v: e.g., motum for movtum, from moveo; (c) by the con- version of V into o (conf. 26): e.g., victum for vivtum, from vivo; fructum for fruvtum, from fruor. 33. Roots in m optionally insert p before the affix t: e.g., demo gives dem-p-tum or demtum: emo, em-p-tum or emtum. 34. Lingual roots in d and t convert the affix t into s, and either drop or assimilate the characteristic: e.g., arsura for ardsum, from ardeo ; cessum for cedsum, from cedo ; versum for vertsum, from verto : messum for metsum, from meto. 35. Lingual roots in r sometimes receive the affix t without modification, but commonly either (a) change the characteristic to s and retain the affix, or (/>) change the affix to s and retain the characteristic. Examples of these three methods of formation are, — pario, partum; gero, gestum; curro, cursum. 36. Final s preceded by a vowel, and medial s between two vowels, is usually changed to r: e.g., amor from amo+s, present indicative passive ; eram for esam ; regerem for regesem. ANALYSIS OF THE VERB. 13 37. A few verbs strengthen some or all of their parts by pre- fixing or alHxing n to the characteristic: e.g., ju-n-go, from root jug; fi-n-do, from root fid; ster-n-o from root ster or stra, 38. A few verbs, mostly inchoatives, are formed by affixing to the root, in the present system, the letters sc with a connective. When formed from verbs, the connective is the same as the con- jugational affix of the primitive. Those derived from adjectives insert e: e.g., inveter-a-sc-o from inveterare; flor-e-sc-o from florere; ingem-i-sc-o from ingemere; obdorm-i-sc-o from obdormire; moU-e-sc-o from mollis. 39. In rare instances the use of sc is not limited to the present system, and in a few cases no connective is employed: e.g., po-sc-o, po-po-sc-i; di-sc-o. In both of these instances the vowel is radical. Vid. " Anomalous Formations." 40. Assimilation. This term, as commonly employed, em- braces a class of euphonic changes, differing not at all in their spirit from another and much larger class, to which no special denomination is applied. We should prefer, therefore, to define assimilation as the accommodation of one letter to the character of a concurrent letter, without at all limiting it to the case in which the assimilated letter becomes a simple reduplicate ; for precisely the same influence which converts s into 1 in vellem (for vel- sem) is also operative in the conversion of g into c in rectum (for reg-tum). 41. As employed in this broader sense, assimilation admits of three cases : (a) in which, under the influence of the assimilating letter, a sonant becomes surd, or a surd sonant ; (6) in which one consonant makes its concurrent letter cognate (i.e., of the same class, either palatal, labial, lingual, or nasal) ; (c) in which the assimilated letter becomes a simple redupficate. Illustrations of these three cases, are, respectively, — ges-tum for ger-tum, ru-m-po for ru-n-po, quas-si for quat-si. j;j"oTE. — Exceptional verbs, not conformable to the rules above stated, will be considered under " Anomalous Formations." Reduplication. 42. Reduplication consists in prefixing to a root its vowel and initial consonant (for the attenuation of the vowel, vid. 11 and 12), and occurs — 14 ANALYSIS OF THE VERB. 43. (A .) In the formation of a few perfects : e.g., tundo (root, tud) becomes, in the perfect, tu-tudi ; mordeo becomes mo- mordi; the root man, to think, makes, in the perfect, memini, to remember. ■44. Remark 1. Two roots, sta (from sto, stare) and spoiid (from spondeo) , prefix the first tivo consonants, dropping, how- ever, the initial from the radical syllable. Thus, ste-ti for ste- sti, spo-pondi for spo-spondi. 45. Remark 2. If the root begins Avith a vowel, the initial only is reduplicated; e.g., ago, perfect a-agi, which becomes (by 12) a-igi, (by 8) egi. 46. The Latin contains a number of concealed reduplicates, whose later form is the result of syncope and contraction; e.g., capio (root, cap) made its perfect originally ca-capi, which be- came successively ca-cipi, ca-ipi, cepi. 47. A few cases occur in which perfects originally reduplicated have lost the initial syllable ; e.g., tiili (from toUo) for the earlier te-tuli, fidi for the earlier fi-fidi. 48. In composition with prepositions, the reduplicate syllable is commonly omitted: e.g., con-tingo, con-tigi ; per-curro makes per-curri, or per-cu-curri. 49. (5.) Rarely in the formation of the present. Examples are: gi-guo (for gi-geno), root gen; si-sto, root sta. CONJUGATIONAL AFFIXES. 50. The conjugational system of the Latin is not the product of Italian soil, but an inheritance received from the progenitor of the Indo-European family of languages, and will be best explained by reference to the Sanskrit, the oldest and fairest representative of that family. Indian grammarians classify Sanskrit verbs into ten conjugations, according to the changes which the roots experience in inflection. Of these, the first, for example, lengthens its radi- cal vowel by combining it with a, and to the root thus strengthened aflixes a, before adding the personal terminations. Thus, budh, to know, by the insertion of a becomes baudh = (by 8) bodh, and, by affixing a, bodha. By uniting -svith this base the termina- tion ti, of the third person singular, there appears the form bod- hati=;/ie knoics. Again, the tenth conjugation lengthens its radical vowel, and to the root thus strengthened affixes aja,* * rronounced as though spelled aya. ANALYSIS OF THE VERB. 15 before adding the personal endings. Thus, kam, to love, becomes successively kam, kam-aja, kam-aja-ti = he loves. This San- skrit tenth conjugation is the archetype of the Latin first, second, and fourth, the affix aja being variously reduced to the forms a, e, i, the characteristics of these conjugations respectively. 51. Remark. This aja is probably an expansion of the root i, to go. Kam-aja-ti may therefore be literally rendered lie goes to love. Compare with this the use of such expressions in the Latin as ire ad gaudium, ire ad cogitationem ; also, the em- ployment of this same verb ire in conjunction with the supine, to form the future infinitive passive ; e.g., amatum iri. Conjugation First. 52. Li this conjugation, the semi-vowel j of the affix aja is syncopated ; and the two a's combine to form a long. The Latin amo has lost an initial palatal, — k, — and is radically identical with the Sanskrit kam, mentioned above. To verify our explana- tion of the a characterizing the first Latin conjugation, we place side by side the inflection of kam and amo, in a part of the pres- ent active indicative forms. Sanskrit. Latin. 2d Sing. kam-aja-si (k)am-a-s(i) 3d Sixg. kam-aja-ti (k)am-a-t(i) 1st Plur. kam-aja-mas (k)am-a-mus oD Plur. kam-aja-nti (k)am-a-nt(i) The only parts of the verb amo, in which, as respects the conjugational affix, the actual seems not to be in accord with the theoretical form, are the first person singular, present indicative active, which will be explained at 123 ; and the present subjunctive, which will be explained at 85. Conjugation Second. 53. Li this conjugation, the second a of the affix aja is elided, the semi-vowel j changed to its corresponding full vowel i, and the resulting form ai passes into e (by 8). That the e of the second conjugation, like the a of the first, is derived from the old affix aja, appears on comparing the inflection of the Sanskrit man with the Latin verb moneo, of the same root. 16 ANALYSIS OF THE VERB. Sanskrit. Latin. 2d Sing, man-aja-si nion-e-s(i) 3d Sing, man-aja-ti nion-e-t(i) 1st Plur. man-aja-mas mon-e-mus 3d Plur. man-aja-nti mon-e-nt(i) The full fonn of the perfect and supine -would be mon-e-vi, mon-e-tum. For the method by which these have been abbre- viated to mon-ui, mon-i-tum, and for the explanation of the first person singular, present indicative active, vid. 125. Conjugation Fourth. 54. Considering next the fourth conjugation (as its affix i has the same origin as the a of the first, and the e of the second) , we find that in this the first a of aja becomes i, the semi-vowel is changed to its corresponding vowel i, and the two i's combine to form i long. The second a is generally dropped : the parts of the verb in which it is retained and changed will be considered at 126. That the i of the Latin is sprung from the Sanskrit aja appears, on comparing the inflection of the Sanskrit svap (conjugated as a causal) with that of the identical Latin root sop. Sanskrit. Latin. 2d Sing, svap-aja-si sop-i-s(i) 3d Sing, svap-aja-ti sop-i-t(i) 1st Plur. svap-aja-mas sop-i-mus 3d Plur. svap-aja-nti sop-iu-nt(i) The third person plural — sop-iu-nt — is one of the forms in which the second a of aja has been retained, and changed to u. Conjugation Third. 55. Two classes of verbs, which may be represented respec- tively by veho and capio, with characteristic vowels distinct in their origin, have been, by grammarians, referred to the third conjugation. 56. The Sanskrit first conjugation affixes to the root simple a, and from this a comes the i, which characterizes verbs like veho ', ANALYSIS OF THE VERB. 17 as, in the forms veh-i-s, veh-i-t, &c. The parts in which a is not affixed, or is changed into another vowel than i, will be considered at 127. For the relation of the Sanskrit a to the Latin i, compare the inflection of vah (first Sanskrit conjugation) , and veho (root, veh) : — Sanskrit. Latin. 2d Sing, vah-a-si veh-i-s(i) 3d Sing, vah-a-ti veh-i-t (i) 1st Plur. vah-a-mas veh-i-mus 3d Plur. vah-a-nti veh-u-ut(i) The third person plural is one of the forms in which a has been changed, not into i, but into u, giving veh-u-nt in place of veh-i-nt. 57. The Sanskrit fourth conjugation adds ja to the root, and to this j a is to be referred the affix of the so-called " Verbs in io, of Conjugation Third." The j of this ja is changed to i,the a is often dropped. The parts of the verb in which it is retained and changed will be considered at 128. We subjoin a specimen of the probable earlier and later inflection of capio : — 2d Sing. cap-ja-si =:cap-i(a)-s =cap-i-s 3d Sing. cap-ja-ti =cap-i(a)-t =cap-i-t 1st Plur. cap-ja-mas = cap-i(a)-inus = cap-i-mus 3d Plur. cap-ja-nti =cap-ia-nt =cap-iu-nt Irregular Verbs. 58. Beside those verbs that are classified into conjugations, there are, in the Latin, a few known as irregular. Irregular verbs are such as add the personal affixes to the root without an inter- mediate vowel, and correspond to the Sanskrit second conjugation. They are of two kinds : — 59. (^.) Irregular verbs, whose root ends in a vowel. These are : do, dare (root, da) ; for, fari (root, fa) ; flo, flare (root, fla) ; no, nare (root, na) ; sto, stare (root, sta) ; in-quam (root, qua); eo, ire (root, i). Of these, all except eo and in- quam are, from their resemblance to the first conjugation, usually classified with that, — improperly, however, for the a of these verbs 2 -\ 18 ANALYSIS OF THE VERB. is not derived from the affix aja, but constitutes a part of fhe root ; and do, for example, is to be analyzed, — not but Root. CONJ. Af. Pers. Af. Root. Peks. Af. 2d Sing. d a s da s oD Sing. d a t da t 1st Plur. d a mus da mus 3d Plur. d a nt da nt 60. (B.) Irregular verbs, whose root ends in a consonant. These are: edo, root ed; volo, root vol; fero, root fer; sum, esse, root es. We subjoin the parallel inflection of the Sanskrit (second conjugation) root ad, to eat, and the identical Latin root ed, giving, in each instance, first the regular, then the euphonic form : — Sais'skrit. Latin. 2d Sing, ad-si (at-si) ed-s (es) 3d Sing, ad-ti (at-ti) ed-t (es-t) 1st Plur. ad-mas ed-mus (ed-(i)-mus) 2d Plur. ad-tha (at-tha) ed-tis (es-tis) By the side of es, est, &c., are found also edis, edit. But it is not to be so understood that es is a contraction of edis : es or eds is the original form, from which, by the insertion of a vowel (i), after the analogy of the conjugations, has been formed the later edis. To the same influence of analogy is to be referred the very general use of a connecting vowel, in forming the parts of fero and volo. Affixes of Voice. 61. The active was the voice earliest developed, and therefore contains, in its inflection, no clement by Avhich it is specially char- acterized. The Latin passive is a modification of the active, and is formed by affixing to the latter the accusative of the reflexive pronoun se, or its initial s. This union is limited to the present system, and is attended with various euphonic changes, as fol- lows : — ANALYSIS OF THE VERB. 19 62. Indicative and Subjunctive. The affix s is changed to r ill all forms ( by 36), except that of the second person singu- lar ; e.g., amo -)- s = amo -]- r = amor. 63. Final m, and final s of the first person plural, of the active forms, are dropped before the appended s : e.g., amabam -|- s = amaba -|- s = amaba -\- r (by 62) = amabar ; audiremus -|- s = audiremu -|- s = audiremu -]-i: = audiremur. 64. Active forms terminating in a consonant, except those men- tioned under 63, affix s with a connecting vowel. This vowel, in the second singular, is i ; in other cases, u; e.g., regeret-}-s = regeret -|- u -|- s = regeret -j- u -[- r = regeretur. 65. In the second person singular, final s of the active becomes r (by 36) ; e.g., capiebas -|- s = (by 64) capiebas -|- i -j- s = capiebar -[- i -|- s = capiebaris. 66. i, immediately preceding s final of the active voice, appears in the passive as e (conf. 65 and 16) : e.g., monebis -|- s = (by 64) monebis -|- i -|- s == (by 65) monebir -]- i -j- s = moneber -|- i -|- s = moneberis. 67. Note. — The second form of the second singular — amare, amabare, &c. — is simply an abbreviation of the regular amiaris, amabaris. 68. Imperative. The only special remark to be made upon the imperative is, that its second person singular adds the full form of the reflexive se to the active voice, with the regular change of s to r : e.g., ama -j- se = ama -\- re ; cape -|- se = cape -|- re. In other respects, this mood is formed according to the rules given above for the indicative. 69. INFI^^TIVE. To form the infinitive, s (changed to r) is added as above, a connecting vowel — e — inserted, and the final e of the active form changed to i. Thus, we have successively, amare, amare-r, amare-er, amari-er. This form amarier, of fi\'(|uent occurrence in earlier Latin authors, is reduced to the later amari, by the ehsion of the final syllable. In the third conjugation, the corruption seems to have been carried still farther, and for forms like regi (obtained by elision from regier) may be conjec- tured an earlier regerier. 70. Remark. We have said that the passive voice is formed from the active by affixing to it the accusative of the reflexive pro- noun, or its initial consonant. Exception to this principle is, how- ever, to be taken in the case of the second person plural. The 20 ANALYSIS OF THE VERB. present indicative — amamini, for instance — cannot be explained as a union of se with the corresponding active amatis. Compare also monebitis with monebimini; regeretis with regeremini. In each instance, there is added to the base of the tense the ending mini ; and this mini, identical in root with the Greek menos, is to be regarded as no other than the nominative plural masculine of a passive participial affix in minus-a-um. Thus, amaminus, amam- ina, amaminum ; plural, amamini, amaminae, amamina. Con- sidered as a participle, this form suits the requirements of the place in respect to number and case, and as usually employed in respect to gender ; for forms like capti estis are much more fre- quent than captae estis, while capta estis would be almost im- possible. Amamini is, then, to be understood as standing for amamini estis ; and the omission of estis must be put upon the same ground as the not infrequent suppression of est and sunt with the existing participle in tus : thus, amatus for amatus est, amati for amati sunt. As confirmatory of the above explanation, it is of interest to notice, in this connection, tra-ces of the use of this same participle in the singidar number; e.g., alumnus (from alo, to cherish) for alu-minus, one that is cherished, hence a fos- ter-child ; ter-minus, /Aa^ which is passed over, from an old Sanskrit root (ter), signifying to go over ; so probably geminus for gen- minus. Modal and Temporal Affixes. We now proceed to review the several tenses of the different moods, considering the affixes, modal or temporal, by which they are respectively characterized. Indicative. 71. Present. This form receives no affix to mark either its mood or tense, the personal endings being added directly to the affix of conjugation; e.g., am-a-t, reg-i-mus. 72. Imperfect. This tense receives a temporal affix identical in its origin with the imperfect of the Sanskrit verb to he. We subjoin the inflection of this Sanskrit imperfect in the singular, omitting the augment. bav-a-m bhav-a-s I ^""'"^^ *^^ ^^'P^'^^" .^' \ bav-a-s bhav-a-t ) r^cn^nfi^.w ( bav-a-t and there remains respectively, ANALYSIS OF THE VERB. 21 These forms are still farther abbreviated by the syncopation of the V, and the union of the two a's, a change identical with that by which aja, the affix of the first conjugation, becomes a, and analogous to that by which the Latin mavolo becomes successively ma-olo, malo. Thus bav-a-m becomes ba-am, bam; bav-a-s becomes ba-as, bas, &c. Of bam, bas, bat, the letters m, s, t, are the respective personal endings : after the elision of these, there remains the common form ba, which is the temporal affix of the Latin imperfect ; e.g. : lOOT. CoNJ. Affix. Temp. Af. Peks. Af. am a ba m = amabam. mon e ba t = monebat. 73. Future. The conjugations are not in hannony in their method of forming the future. The affix of the third and fourth is strictly a modal one, and will be considered with the present subjunctive. That of the first and second is temporal, and kindred to the ba of the imperfect. The root of the Sanskrit verb to be is bhu, or, dropping the aspiration, bu. If to this root deprived of its vowel be added the Sanskrit future character ja, there results the form bja. We have already seen (vid. 57) that the letters ja, with the personal affixes, added to the root cap give the forms, — Singular. Plural. cap-io- cap- i -mus cap- i -s cap- i -tis cap- i -t cap-iu-nt The inflection of b-ja we should therefore expect would give, b-io- b- i -mus b- i -3 b- i -tis b- i -t b-iu-nt After syncopating the first of the two vowels in the first singu- lar and third plural, and dropping the personal endings, there 22 ANALYSIS OF THE VERB. remains the common form bi (first singular, bo ; third plural, bu), which is the affix of the future indicative ; e.g. : Root. CONJ. Af. Temp. Af. Pers. Af. am a bi mus = amabimus. mon G bi tis = monebitis. 74. Perfect. In considering the temporal sign of the per- fect, verbs may be divided into three classes. To the first class belong those which form the tense by reduplication (vid. 42-49). They are the following : — ago * frango * posco cado • * jacio * pungo caedo * memini * scindo cauo mordeo * sisto capio * pango spondeo cello (obsolete) parco * sto curro pario * tango disco pedo * tendo do * pello tondeo facio pendeo * tollo faUo pendo * tnudo findo 75. To the second class belong verbs which employ in the per- fect the affix u (after vowels convertible into v) . The explana- tion of this letter is found in the perfect of the same Sanskrit root bhu, to he, whose imperfect and future have furnished respec- tively the affixes ba and bi. The base of the Sanskrit perfect is ba-bhuv. If from this form there be dropped the reduplica- tion (conf. fidi from an earlier fi-fidi; tiili from an earlier te-tuli), there remains bhuv. Now the Sanskrit bh is the equivalent of the Latin f. Conf. Sanskrit bhar, to carrj/, with fer-o: bhuj, to turn, vriih fug-io. With the substitution of f for bh, bhuv becomes fuv, or, with the connecting vowel i, fuvi, — a form actually occurring in the earher Latin authors, as the perfect of sum. From this by syncopation is derived the * Vid. " Auonialous Formations." ANALYSIS OF THE VERB. 23 later fui. In the word pos-sum for pot-sum, pot (for potis, able) is purely an adjective element ; hence, in the perfect pot-iii, the verbal force must reside -wholly in .the ui. As the present of possum is compounded of potis and the present of sum, we should have expected, as the perfect of the same verb, pot-fui; i.e., potis -|- the perfect of sum. As then ui is verbal in its character, as the actual pot-ui so far coincides with the conjectural pot-fui, and as there is no other primitive with which ui admits of being associated, we shall be justified in identifying it with fui, and in recognizing this u (i is simply connective) as the old ba-bhuv carried to the utmost limit of abbreviation. To the same origin must of course be referred the u (or v) characterizing the perfect tense of other verbs ; e.g. : Root. Conj. Ap. Temp. Af Con. Vow. Pees. Ap. am a V i t = amavit mon (wanting) u i t := monuit. aud i V i mus = audivimus. 76. To the tliird class belong verbs which in the perfect add to the root the affix s. The Latin verb to he emplo}S in its conjugation tAvo roots ; fu, already considered, and es (Sanskrit as). From the latter is formed, among other tenses, the imperfect esam (later eram, by 36). Identical with esam is the Sanskrit asam. Now to form the tense corresponding to the Latin perfect in s, the Sanskrit adds to the root the several numbers and persons of asam. It may therefore with consid- erable confidence be conjectured that the Latin esam originally rendered the same service, and that to this, as its origin, is to be referred the affix s under consideration. Examples of perfects in s are : — Root. Temp. Af. Con. Vow. Pers. Ap. reg si t = rexit. nub s i mus = (by 25) nupsimus. 77. Instances occur in which the last two modes of designating the perfect are combined: e.g., met-f s + u-|-i= (by 28, c) messui ; nect -f s + u -f i = (by 28, a) iiec -f s + u + i = (by 23) nexui. 24 ANALYSIS OF THE VERB. 78. Aside from the three classes specified above, there remains a large number from which all trace of an original prefix or affix has disappeared, save, perhaps, in the changed quantity of the radical vowel, — e.g., vidi from root vid ; legi from root leg, — where the lengthening of the vowel may be regarded as compen- sation for the syncopation of a consonant, or the aphaeresis * of the reduplicate syllable. It is to be supposed that all Latin verbs, certainly all which had an existence in the earlier period of the language, originally designated the perfect by some special char- acteristic. In its loss we only see carried one step farther the process of corruption by which ba-bhuv has been reduced to u, and esam to s. 79. Pluperfect. This tense is formed by adding to the base of the perfect, the imperfect of the verb sum. Thus rex (for reg-f s)-erat, amav-eram, ce-cin-eratis. In the first of these examples, the root reg expresses abstractly the idea of "ruling." The form reg -\- s expresses concretely the idea of ' ' having ruled ; " and the form reg -f- s + era -f- 1, with its double temporal and single personal limitation, reduces the generic conception " rul- ing " to the specific thought " he was having ruled," i.e., " he had ruled." 80. Future Perfect. The future perfect adds to the base of the perfect the future of sum; e.g., amav-ero, rex-ero, ce-cin-ero. 81. There occurs also an old future perfect in so, formed by adding to the base of the perfect eso (later ero; vid. "Table I."), with the aphaeresis of e: e.g., amas-so, by assimilation for amav-so : habes-sit for habev-sit ; cap-so, probably for ca- cap-so; faxo (fac-so) for fa-faxo. Vid. capio and facio under "Anomalous Formations." Subjunctive. 82. Present. The Latin Subjunctive, though known as a single mood, in reality comprises two, which are distinct in origin, and may be designated the Conjunctive and Optative. 83. To the conjunctive belongs the present subjunctive of con- jugations second, third, and fourth. Its modal affix is a. * Aphseresis is the loss of a syllable or letter at the beginning of a word. ANALYSIS OF THE VERB. 25 The full form of moneo, in the present subjunctive, first plural, will therefore be : — Root. Conj. Ap. Mod. Af. Pers. Af. mon aja a mus or (changing aj to e by 53), mon ea a mus or (combining the a's of the conjugational and modal affixes into a long) m,on e a mus = moneamus. The full form of rego, present subjunctive, first plural, would in like manner be : — Root. Conj. Af. Mod. Af. PeFvS. Af. reg a a mus or (combining the two a's into a long) , reg a mus == regamus. In the same manner, the present subjunctive, first person plural, of audio, would be : — Root. Conj. Af. Mod. Af. Pers. Af. aud aja a mus or (changing aj into i by 54), aud ia a mus or (combining the a's of the conjugational and modal affixes into a long), aud i a mus=:audiamus. 84. As conjunctives are to be reckoned, also, such forms as creduam, perduam, in which a is the modal affix, and u the attenuation of the radical vowel of the root da. Vid. do under " Anomalous Formations ; " also 59. 85. To the optative belongs the present subjunctive of conjuga- tion first, and the future indicative of conjugations third and 26 ANALYSIS OF THE VERB. fourth. The modal affix is ja. The fall form of amo, in the present subjunctive, third person plural, would therefore be : — Root. Conj. Af. Mod. Af. Pers. Af. am aja ja nt or (converting aja into a by 52, and dropping the a of ja), am a j nt or (changing j to i by 9, and combining ai to form e by 8), am e nt =: ament. By a like process is obtained the later form of rego, future indicative, third plural. Root. Conj. Af. Mod. Af. Pers. Af. reg a ja nt or (by the elision of the modal a, the conversion of j into i, and the union of a with i) , reg e nt = regent The full form of audio, future indicative, third plural, would in like manner be : — Root. Conj. Af. Mod. Af. Pers. Af. aud aja ja nt or (by the change of aj to i, by 54, and the conversion of the modal j to i), aud ia ia nt. or (by the ehsion of the modal a, and the union of the conjuga- tional a with the modal i to form e), aud i e nt = audient The first person singular of these futures is, however, to be ex- plained as conjunctive rather than optative, and hence forms regam for regem, and audiam for audiem. ANALYSIS OF THE VERB. 27 86. As optatives are to be reckoned also forms like du-i-m, credu-i-m, perdu-i-m (conf. 84) ; also ed-i-m, to eat (conf. 60). 87. Impekfect. The verb is conjugated in this tense by the use of the imperfect subjunctive forms of the verb to he, sem, ses, set, &c. (vid. "Table I."), "vvhich after the conjugational affix or a connecting vowel become (by 36) rem, res, ret. Ex- amples are : — Root. Conj. Af. Ending. am a rem = amarem. mon e ret = moneret. reg e remus = regeremus. aud i rent = audirent. 88. Perfect. This tense adds to the base of the perfect indicative, the present subjunctive of the verb to he, with the connecting vowel i. Hence, for example, results the form, — Root. Conj. Af. Temp. Af. Con. Vow. Ending. am a V i sim. or — with the change of s to r (by 36), and the displacement of i by e before r (by 16) — am a V e rim = amaverim. 89. By the omission of the connective, and consequent reten- tion of the s, are obtained such forms as negas-sim, by assimila- tion for negav-sim ; emis-sim for emit-sim ; faxim for fac-sim, and this for a conjectural fa-fac-sim. Conf. 81. 90. Pluperfect. This tense joins to the base of the perfect indicative, by means of a connective i, the imperfect subjunctive forms of the verb to be, sem, ses, set, &c., with the reduplication of the initial consonant s; e.g. : Root. Conj. Af. Temp. Af. am. a v reg (wanting) s The reduplication of the s, though of rare occurrence in the Latin, is not without its analogies in the Greek. The i is to be considered purely as a connective, and in no sense an attenuation of the e of essem: i before two consonants frequently becomes Con. Vow. Ending. i ssem = amavissem. i ssem ^^ rexissem. 28 ANALYSIS OF THE VERB. e, but not the converse. Amavessem might have proceeded from amavissem, but not amavissem from amavessem. Imperative. 91. The imperative receives no modal or temporal affix, but adds the personal endings directly to the affix of conjugation. Infinitive. 92. The Latin infinitives are simply oblique cases (dative, accusative, or ablative) of the noun. As a dative is to be consid- ered, j^?-5^, the infinitive present. This adds to the root, with its conjugational affix, the ending se (later re) . Root. Conj. Af. Mod. Af. am a se :^ amare. This ending is still farther to be resolved into s, the radical consonant of the verb to he, and e, the case affix of the dative. The use of this e as a case sign appears, for example, in such datives as causae = causa -|- e ; aquee == aqua -\- e. The i of forms like Romano-i, populo-i, &c. (later Romano, populo) is the second element of this e (= a -|- i) • 93. Second, the infinitive perfect active. This tense joins to the base of the perfect indicative, by means of the connecting vowel i, the ending se with the reduplication of its initial s (conf. 90) ; e.g. : Root. CONJ. Ap. Temp. Ap. Con. Vow. Mod. Ap. aud i V i sse. 94. TJiird, the infinitive perfect passive. This tense compounds the perfect passive participle with the infinitive esse. For the explanation of esse, which is also a dative, vid. 134. 95. Fourth, the future infinitive active, which combines the future active participle with esse. 96. Remark 1. The relations most commonly denoted by Latin datives are those expressed by the prepositions to and jfoi'. Precisely analogous, then, is the use of dative infinitives in Latin to that of the English infinitive in the objective case governed by to: e.g., to be, i.e., to the being; to love, i.e., to the loving. As expressing the proper case relation of infinitives in e, the sen- ANALYMS OF THfe , YEB^ 29 tences solent cogitare, non audeo confiteri, may be respectively rendered, they are accustomed to thought, I have not the courage for a confession. 97. Remark 2. That infinitives strictly dative should be em- ployed to denote a variety of relations foreign to the dative, is but another illustration of the principle, that the exact use of forms can exist only with the consciousness of their exact import. So long as e was remembered as a case affix, and to recognized as a preposition, amare, and to love could only render the service of a dative and objective ; but the gradual identification of these elements with the body of the word resulted also in the loss of those definite relations of which they had been the symbol. 98. As an accusative is to be considered the infinitive com- monly known as the "supine in m.'" This adds either to the conjugational affix, or to the root with or without the connective i, the affix tiim. Examples of these three modes of attachment are, — am-a-tum, mon-i-tum, dic-tum. For the correspond- ence of the Sanskrit infinitive with this Latin form compare San- skrit stha-tum, to stand, with sta-tum; pa-tum, to drink, with po-tum; jan-i-tum, to beget, with gen-i-tum; da- turn, to give, with da-tum. This affix may be resolved thus, tii-m. Tii is an element quite commonly employed in the Latin to form nouns of action and condition. Conf sta-tu-s, tac-tu-s, son-i-tu-s, or-tu-s. m is the case sign of the accusative, common to all the declensions. This infinitive is used after verbs of motion, to denote that in which the motion ends, and hence as the oblique case of a substantive falls readily under the Accusative of Limit. 99. In the "supine in u" may be recognized an ablative infinitive. This is formed from the same base as the infinitive in m, and is characterized by the affix tu (for tu-(d). This ending may be resolved thus, tu-u. The first element is identical with the tu of tii-m, considered above : the u final is a reduplicate of the radical vowel (received in the ablative for emphasis), and combines with it to form u long. Compare the nominative fruc- tu-s with the ablative fruc-tu (=fruc-tu-u), for the earlier fruc- tu-d. This infinitive is employed to define the application of adjectives, and hence, as the oblique case of a substantive, is to be classed with ablatives of specification. 100. Remark 1. In the future passive this mood combines the dative infinitive iri (impersonal passive from eo) with the accusa- 30 ANALYSIS OF THE VERB. tive infinitive in m. In the use of this accusative after eo, note the conformity with the ^^rinciple stated at the close of the 98th section. 101. Remark 2. Tliough properly the modal affix of these three varieties of infinitive is respectively se, turn, and tu, for the sake of convenience the formative case elements, e, m, and ii, will in the "Tables" be separated from the base, and entered under "Case Affix." Participle. 102. Tlie Latin participles are formed in the use of but three distinct affixes. First, nt, which is added to the affix of conjuga- tion to form the participle present. To the base as thus composed there is affixed in the nominative the case sign s ; e.g. : Root. Conj. Ap. Mod. Ap. Case Ap. am a nt s = (by 28, a) amans. 103. With the exchange of the surd lingual t for the sonant lingual d, nt, the affix of the verbal adjective (participle) is converted into nd, the affix of the verbal noun (gerund) ; e.g. : Root. Conj. Ap. Mod. Ap. Case Ap. mon e nd i (genitive) 104. nt with the addition of u, and the conversion of t into d, as above, forms the affix of the future participle passive ; e.g. : Root. Conj. Ap. Mod. Ap. Case Ap. am a ndu s (nominative) 105. Any reluctance to identify such forms as amans and amandus, on the ground that one is passive, and the other active, may be removed by noting the traces of an original active force in both. Thus from sequor, to follow ; voluto, to roll ; and vito, to avoid, — come respectively : — Root. Conj. Ap. Mod. Ap. Case Ap. sequ — * ndu s = secundus, the following one. volut a (bu) ndu s = volutabundus, tlte rolling one. vit a (bu) ndu s = vitabundus, the avoiding one. * A dash deaotes that a part is wanting. ANALYSIS OF THE VERB. 31 The last example is particularly interesting, in that it is not only used in an active sense, but is also employed by Livy with a dependent accusative : Hanno vitabundus castra hostium. 106. IsoTE. — The letters bu in forms like vita-bu-ndus, treme-bu-udus, are referred to the root bhu (Latin fu, fu-i ; vid. also 72 and 73). 107. The second variety of participial affix is turu, which is added to the affix of conjugation, or to the root Avith or without the connective i, to form the future participle active ; e.g. : Root. Con J. Af. Con. Vow. Mod. Af. Case Af. am a — turu s mon — i turu s die — — turu s 108. u final of the ending tur-u is inorganic (conf. at 104 nd-u from nt) . Other forms of the affix tur are tor and tri. The former of these is combined with roots to form masculine nouns of agency, reckoned by Sanskrit grammarians as one of the forms of the verb. Examples are, gen-i-tor, vic-tor, sta-tor. The latter is added to roots to form feminine nouns of agency, with, however, the insertion of c before the case endings : e.g., gen-i-tri-c-s = genitrix ; vic-tri-c-s = victrix. The Sanskrit root tar is probably the primitive of the later tur, tor, tri, and signi- fies a doer or maker. Both da-tor and da-tur(u)-s therefore literally signify one making a gift. In the development of the language, the action of the latter became limited to the future, while that of the former remained undetermined. 109. The third form of participial ending is tu, which is added to the conjugational affix, or to the root with or without the con- nective i, to form the perfect participle passive ; e.g. : Root. CoNj. Af. Mod. Af. Case Af. aud i tu s die tu s Personal Affixes. 110. The personal affixes are the effete representatives of old personal pronouns. We proceed to pass in review the sevei-al 32 ANALYSIS OF THE VERB. endings, singular and plural, giving to each that which is its most probable and best approved derivation. 111. First Person Singular. The old root of the first per- sonal pronoun is ma. This root, with, however, the attenuation of its vowel, appears in the oblique forms me-i, mi-hi, me-. As still farther reduced to m, by the surrender of its voAvel, it con- stitutes the affix of the first person. In but two verbs, su-m (compai^ the English a-m) , and inqua-m, is this m retained in the present indicative. It has also disappeared from the perfect and future perfect indicative, and from the future of conjugations first and second. Examples of the use and abandonment of this aflix are respectively : — Root. Conj. Af. Mod. or Temp. Af. Pers. Af. mon e a (Mod) m am a bo (Temp.) — 112. Second Person Singular. Tva, the root of the second personal pronoun, by the loss of a and the conversion of v into u, has been corrupted to the Latin nominative tu. The same root (tva) has become successively ta (by the loss of v) , ti (by attenuation), si (by interchange betAveen lingual surds), s by elision of the vowel. In the form of s it constitutes the affix of the second person. 113. In the imperative the ending has disappeared : in the perfect indicative it exists in the earlier form ti, before Avhich is inserted a euphonic s (for the insertion of s, compare such Greek forms as ke-keleu-s-tai for the regular ke-keleu-tai) . Exam- ples of the second person are : — Root. CoNj. Af. Temp. Af. Con. Vow. Pers. Af. am a ba — s aud i V i (s) ti mon e — — — 114. Third Person Singular. Ta, a Sanskrit pronominal root of the third person, becomes by attenuation to or te (conf. I^atin is-te) , or, by the surrender of its vowel, t. In this latter form it serves as the third personal affiux. In the imperative alone is retained the earlier form, to. ANALYSIS OF THE VERB. 33 115. Remark 1. The third person in to may also perform the office of the second ; hence am-a-, or am-a-to. 116. Remark 2. An old form of affix in the imperative is tod. Thus in the ' ' Twelve Tables " occurs the expression Fidios af patre leiber estod; i.e., Filius a patre liber esto. This ending is to be explained as the reduplication of to with suppres- sion of the second vowel, and the conversion of t into its cognate sonant. Examples of the third person are : — Root. CONJ. Af. Mod. Af, Pers. Af aud i a t aud i — to 117. First Person Plural. Mus, the affix of this person, is a derivative by elision and attenuation from the primitive masi ; i.e., ma-J- si. Of tliis compound the element ma is identical with the pronominal root of the first person singular. The second element si is one of the forms which we have seen to be derived from tva, the pronominal root of the second person singu- lar; ma + si therefore = 1 + you = we. This person affords no exceptional cases. 118. Second Person Plural. Tis, the affix of the second plural, is derived by elision from tisi, and this by attenuation from tasi = ta -}- si. The first element, ta, is derived from the pronominal root of the second person singular (vid. 112) ; and si, the second element, is a later derivative from the same root. ta -|- si therefore ^^ you (sing.) -[- you (sing.) = you (plural) . For the insertion of s in the perfect indicative, vid. 113. 119. To form the imperative, the root ta is reduplicated to ta-ta, and the vowels attenuated respectively to o and e, hence to-te. Examples of the second person are : — Root. CONJ. Af. Temp. Af. Con. Vow. Pers. Af. mon e bi — tis aud i V i (s) tis am a — — to-te 120. Third Person Plural. The full form of ending in this person was probably anta, — i.e. , aii -}- ta, — of which an is an old demonstrative root, and ta identical with the pronominal root of the third singular. An -\- ta therefore = he -j- he = they. 3 34 ANALYSIS OF THE VERB. 121. The first a of anta is retained only wlien the ending is attached to consonantal bases, and is then modified to u, under the influence of the nasal n (by 15). The second a appears only in the imperative, and that under the attenuated form of o. Exam- ples of the third person are : — Root. COKJ. Ap. Temp. Af. Pers. Af. am a ba lit mon c — nto (e)s — — unt = sunt 122. Perfects indicative, like amaverunt, require still farther explanation. This form is the corruption of a regular am-a-v-i-iit. Into this, after the analogy of the second persons of the same tense, was introduced an s, after which the ending was added in the form of unt (by 121). Am-a-v-i-s-unt, as thus obtained, was still far- ther modified by the conversion of s to r (by 36) and the change of the preceding i to e (by 16), Hence, amaverunt, or, in its abbreviated form, amavere. Irregularities of the Conjugational Affix. The readiest apprehension of these irregularities seemed de- pendent upon a prior acquaintance with modal, temporal, and per- sonal affixes. The discussion of them has therefore been deferred to the present time. Conjugation First. 123. In the first person singular of the present indicative, the aflSx aja becomes successively aa (by syncopation) , ao (by attenu- ation), o (by contraction). Hence, am-aja, am-aa, am-ao, am-o. A few verbs of this conjugation omit the affix in the per- fect and supine systems, ordinarily inserting in the latter, however, the connective i: e.g., dom-o, dom-ui, dom-(i)-tum; but sec-o, sec-ui, sec-tum. 124. Note. — The i regularly occurring in the perfect of the indicative and infinitive and the perfect and pluperfect subjunc- tive, of all conjugations, we have also preferred to distinguish from the affix, and to designate as connective. ANALYSIS OF THE VERB. 35 CoN.juGATiox Second. 125. In the fii-st person singular of the present indicative the final a of aja is retained, and attenuated to o. Hence, mon-aja becomes successively mon-aia, mon-ea, mon-eo. In the present subjunctive, it is likewise preserved, and. combines with the modal a to form a long. A few verbs like fleo, fle-vi, fle-tum extend the use of the affix to the three systems ; but it is. in almost all cases, limited, to the first: e.g., doc-eo, doc-ui, doc-tnim ; aug- eo, aiisi. auc-tum. In the third system, a connective i is often insened : e.g.. mon-eo, mon-ui, mon-(^i)-tum ; hab-eo, hab-ui, liab-(i)-tuiiL CoNJTGATiox Fourth. 126. The a final of the affix aja is retained and modified in the following forms : it is attenuated to o in the first person singular of the present indicative ; to u in the third person plural of the same tense, as also in the third plural of the imperative ; to e in the imperfect indicative, and before the participial affix nt ( nd, ndu). Examples of the above changes are respectively, aud-io, aud-iu-nt, aud-iu-nto, aud-ie-bant, aud-ie-n(t)s. In the con- jimctive forms, the final vowel of the affix combines with the modal a to form a long. In the optative, it unites with the modal i to form e; e.s. : ElCK)T. Co>-.J. At. Mod. Af. Pess. Af. aud la a mus — audiamus aud ia i(a) nt = audient The use of the conjugational affix is, in some instances, confined to the present system: e.g., aper-io. aper-ui, aper-tum; vinc-io, vinc-si (= vinxi), vino-tum. COX-HTGATIOX ThIRD. 127. We observe, here, the method of classification adopted at 55-57. The parts of the verb which, in conjugation fourth, re- tain, in its modified or unmodified form, the final vowel of the aflix aja, treat in the same manner the affix a in the^r.?^ class of conju- gation third. In the second singular imperative, i is converted to 36 ANALYSIS OF THE VERB. e (by 18), except in the case of the verbs dico, duco, and fero, where it is dropped. In the imperfect subjunctive and present in- finitive, i becomes e under the influence of r (according to 16). 128. The parts of the verb which, in conjugation fourth, retain, in its modified or unmodified foi-m, the final vowel of the affix aja, retain and treat in like manner the final vowel of the affix ja in the second class of conjugation third, i is converted to e in the same forms of the verb, and in obedience to the same principles as in class first, except in the second singular imperative of facio, where it is dropped. This conjugation limits the use of its affixes a and ja to the present system, — in rare instances, however, in- serting in the third a connective i: e.g., claud-o, clau(d)-si, clau(d)-sum; but gem-o, gem-ui, gem- (i) -turn. TABLE I. Conjugation of Esse. The parts of the Latin verb to^ he are formed in the use of two roots, — es and fu. Although s is rarely found in the parts formed from the first of these, yet that it was originally radical appears from the comparison of the third singular, present indicative, in the cognate languages : — Sanskrit. Zend. Lithuanian. Greek. Latin. Gothic. as-ti as-ti es-ti es-ti INDICATIVE. Present. es-t is-t SINGULAR. PLURAL. Probable Primititb Later Derivative. Probabls Primitive. Latkr Derivative. 1. es-mi s-u-m es-mus s-u-mus 2. es-si e-s es-tis es-tis 3. es-t es-t es-ant Imperfect. s-uut 1. es-a-m er-a-m es-a-mus er-a-mus 2. es-a-s er-a-3 es-a-tis er-a-tis 3. es-a-t er-a-t es-a-nt er-a-nt ANALYSIS OF THE VERB. 37 SINGULAR. Probable Peimitive. Later Debivatite. 1. es-io-m er-o- 2. es-i -s er-i-s 3. es- i -t er- i - 1 Future. PLURAL. Pbobable Primititi. Later Debivatitb. es-i -mu3 er-i -mus es- i -tis er-i -tis es-iu-nt er-u-nt Perfect. 1. 2. 3. fu-i-m fu-i-ti fu-i-t fu-i- fu-i-mus fu-i-(s)ti fu-i-tis fu-i-t fu-i.nt Pluperfect. fu-i -mus fu-i-(s)tis fu-e-(r)unt 1. 2. 3. fu-esam fu-esas fu-esat fu-eram fu-esamus fu-eras fu-esatis fu-erat fu-esant fu-eramus fu-eratis fu-erant Future Perfect. 1. 2. 3. fu-esiom fu-esis fu-esit fu-ero fu-eris fu-erit fu-esimus fu-esitis fu-esiunt fu-erimus fu-eritis fu-erint SUBJUNCTIVE. Present. 1. 2. 3. es-ja-m es-ja-s es-ja-t s-i-m s-i-s s-i-t es-ja-mus es-ja-tis es-ja-nt Imperfect. s-i-mus s-i-tis s-i-nt 1. 2. 3. es-a-i-m es-a-i-s es-a-i-t es-(s)-e-m es-a-i-mus es-(s)-e-3 es-a-i-tis es-(s)-e-t es-a-i-nt es-(s)-e-mus e3-(3)-e-tis es-(s)-e-nt Perfect. 1. 2. 3. fu-i-sim fu-i-sis fu-i-sit fu-e-rim fu-e-ris fu-e-rit fu-i-simus fu-i-sitis fu-i-sint fu-e-rimus fu-e-ritis fu-e-rint 38 ANALYSIS OF THE VERB. Pluperfect, singular. plural. Probable Pbimitite. Later Derivative. Probable Primitive. Later Derivative, 1. fu-i-sem fu-i-(s)-sem fu-i-semus fu-i-(s)-semus 2. fu-i-ses fu-i-(s)-ses fu-i-setis fu-i-(s)-setis 3. fu-i-set fu-i-(s)-set fu-i-sent fu-i-(s)-sent IMPERATIVE. 2. es-ti, or es-, or es-te, or es-te, or es-to es-to es-to-te es-to-te 3. es-to es-to es-anto INFINITIVE. s-unto Present. Perfect. es-e es-(s)-e fu-i-se fu-i-(s)-se Future, futurus esse. PARTICIPLE. Present. Future. es-e-nt-s s-e-n-s. fu-turu-s. Remarks on Esse. 129. Indicative. Present. The comparison of such forms as the Sanskrit as-mi, Zend ah-mi, Greek (Doric) em-mi (for es-mi) , Lithuanian • es-mi, points to the existence of an original Latin es-mi, the final vowel of which is the attenuated representa- tive of the old a of the pronominal root ma. On the disappear- ance of the vowel, the consonant was appended in the use of a connective u, which connective appears also in the first person plural, s-u-mus for s-mus (conf. vol-u-mus for vol-mus). A comparison of the same languages renders probable also an origi- nal es-si in the second singular. In this person, however, after the disappearance of the final vowel, the verb preferred the sacri- fice of a consonant to the insertion of a connective, and thus reduced the intermediate es-s (possible only in theory) , to the ANALYSIS OF THE VERB. 39 later e-s. We prefer to regard the syncopation as removing the s of the root rather than of the affix, as of the two consonants of an hnpossible combination the syncopation usually falls upon the first. 130. Imperfect. This tense inserts a connective a throughout, and makes the usual change of s to r between two vowels. 131. Future. The future es-io-m (or es-io — , later es-o) es-i-s, es-i-t, &c., is formed from the root es in the use of the same affix ja, by which b-io (later b-o) b-i-s, b-i-t is produced from the radical b (vid. 73). It requires to be noticed that while the verb, in its simple form, retains in the third plural the latter of the two vowels iu, when compounded with the base of the perfect to form the future perfect, the former only is preserved. Hence fu-er-iu-nt becomes, not fu-er-u-nt, but fu-er-i-nt; am-a-v-er- iu-nt gives, not am-a-v-er-u-nt, but ani-a-v-er-i-nt. 132. Subjunctive. Present. This form of esse is optative, and affixes as its modal sign ja, which becomes by the usual con- version ia, and by attenuation ie. Hence the earlier forms s-ie-m, s-ie-s, s-ie-t, &c. In later Latin ie are contracted into long i. Thus, s-i-mus, s-i-tis. In explanation of the short i of s-i-m and s-i-t, vid. 17. The present subjunctive fuam, fuas, &c., is to be explained as an old conjunctive form, based upon the root fu, and appending the conjunctive modal affix a ; hence, fu-a-m, fu-a-s. 133. Imperfect. This tense is optative, and is formed from es-a, the base of the imperfect indicative, by affixing i of the modal sign ia. Hence, es-a-i-tis, es-a-i-nt, &c., which, by the union of ai to form e, become respectively es-e-tis, es-e-nt, forms of actual occurrence in earlier authors. For the derivation of es(s)-e-m, &c., from the earlier es-e-ra, conf. 90. The forms es-e-m, es-e-s, &c., with the apha3resis of the vowel, constitute the affix of the several conjugations in the imperfect subjunctive ; e.g., am-a-rem, reg-e-rem for am-a-sem, reg-e-sem. Forem, probably a corruption of fu-rem, is to be explained as a union of the same ending rem (for sem) with the root fu. 134. Infinitive. Present. The present infinitive adds to the root, es, the case sign of the dative, e. This form es-e, which occurs in earlier Latin, developed by the reduplication of its consonant into the later e3(s)-e. Fore, another form of the same tense, with however, in general, a future signification, stands 40 ANALYSIS OF THE VERB. for fure, and tliis forfu-se; i.e., the sign of the dative infinitive : added to the root fu. i 135, Participle. Present. This participle appears only in ; composition, and then with the aphaeresis of e. Hence, nomina- \ tive ab-sens (for ab-s-e-nt-s) , genitive ab-s-e-nt-is. The ! compounds absens and praesens have been commonly resolved ; thus : abs-ens, praes-ens. The s belongs, however, in each • case, to the second element, and represents the old root es. . ANALYSIS OF THE VERB. 41 TABLE II. Conjugation of Amo. Active Voice. ANALYTIC. SYNTHETIC. Mood and Pars. Koot. CONJ. Ar. Mod. and Con. Pers. Affix. Tense. Early. Late. Temp. Af. Vow. S. 1. am (aa)* ao _ _ _ amo. |l 2. am (aa) a — — S amas. 3. am. (aa) at — — t amat. ll P.l. am (aa) a mus amamus. 2. am (aa) a — — tis amatis. M 3. am (aa) a — — nt amant. S.l. am. (aa) a ba _ m amabam. II 2. am (aa) a ba — s amabas. 3. am (aa) a ba t amabat. II P.l. n.m (aa) a ba — mus amabamus. 2. am (aa) a ba — tis amabatis. ^° 3. am. (aa) a ba — nt amabant. S.l. am (aa) a bo amabo. ot 2. am (aa) a bi — s amabis. S Is 3. am (aa) a bi — t amabit. |.i P.l. am (aa) a bi — mus amabimus. ^1 2. am (aa) a bi — tis amabitis. 3. am (aa) a bu — nt amabunt. S.l. ara (aa) a V ] amavi. ^i 2. am (aa) a V (s)$ ti amavisti. «2 ^ 3. am (aa) a V } t amavit. P.l. am (aa) a V ] mus amavimus. &i 2. am (aa) a V ] (s)t tis amavistis. M 3. am (aa) a V i nt amaverunt. § S. 1. am (aa) a v-era _ m amaveram. 1^ 2. am (aa) a v-era — s araaveras. *§■! 3. am (aa) a v-era — t amaverat. ti P.l. am (aa) a v-era — mus amaveramus. s-^ 2. am (aa) a v-era tis amaveratis. Sm 3. am (aa) a v-era — nt amaverant. ^ S.l. am (aa) a v-ero _ amavero. %t 2. am (aa) a v-eri — s amaveris. (2-^ 3. am (aa) a v-eri — t amaverit. II gh-l P.l. am (aa) a v-eri — mus amaverimus. 2. am (aa) a v-eri — tis amaveritis. 3. am (aa) a v-eri — nt amaverint. * In this conjugation we carry back the conjugational affix no farther than to aa, the form resulting from aja by the syncopation of j. t Shortened from a long by the influence of the following t final (Vid. 17 )• t Vid. 113 and 118. § Yid. 122. 42 ANALYSIS OF THE VERB. ANALYTIC. SYNTHETIC. Mood and Pers. Root. CONJ. Af. Mod. and Con. Pers. Affix. Tense. Early Late. Temp. Af. Vow. i. S. 1. am (aa) a i(a)* m amem. 11 2. am (aa) a i(a) — s ames. 3. am (aa) a i(a) — t amet. s ^ P.l. am (aa) a i(a) — mus amemus. f^f 2. am (aa) a i(a) — tis ametis. 3. am (aa) a i(a) — nt ament. _ ej S.l. am (aa) a re _ m amarem. t-S 2. am (aa) a re — s amares. fe fl 3. am (aa) a re — t amaret. S.3 P.l am (aa) a re — mus amaremus. ^? 2. am (aa) a re — tis amaretis. M 3 02 3. am (aa) a re — nt amarent. S S.l. am (aa) a v-eri m amaverim. 2. ara (aa) a v-eri — s amaveris. 1« 3. am (aa) a v-eri — t amaverit. P.l. am (aa) a v-eri mus amaverimus. (2;? 2. am (aa) a v-eri — * tis amaveritis. ^ 3. am (aa) a v-eri — nt amaverint. f Mod. and > 1 Temp. 1 V S.l. am (aa) a V sse m amavissem. 1-5 2. am (aa) a V sse s amavisses. "SI 3. am (aa) a V sse t amavisset. P.l. am (aa) a V sse mus amavissemus. PS" 2. am (aa) a V sse tis amavissetis. s^ 3. am (aa) a V sse nt amavissent. aj S. 2. am (aa) a ama. .fe 2. am. (aa) a — — to amato. 1 3. am (aa) a — — to amato. P. 2. am (aa) a — te amate. tt 2. am (aa) a — — tote amatote. 3. am (aa) a — — nto amanto. Mood and Tense. Case Affix. Pres. Inf. am (aa) a r — e t (Dat. ) amare. Perf. Inf. am (aa) a V i ((SSM. Af.)( 1 e (Dat.) ) amavisse. Former Supine. am (aa) a tQ — m (Ace.) amatum.. Latter Supine. am (aa) a tu — u(Abl ) amatu. Pres. Part. am (aa) a n(t) — S (Nom.) amans. Gerund. am (aa) a nd — i^Gen.) amandi. Fut. Part. (Pas.) am (aa) a ndu — S(Nom.) amandus. Perf. Part. (Pas.) am (aa) a tu — S (Nom.) amatus. Fut. Part. am (aa) a turu — S(Nom.) amaturus. Fut. Inf. t * Brackets are here used to enclose that element of the modal affix which does not appear in the synthetic form, t Vid. 101. t The future infinitive combines the future participle Mith esse. ANALYSIS OP THE VERB. 43 TABLE III. Conjugation of Moneo. Active Voice. Present System. ANALYTIC. SYNTHETIC. Mood and Pers. Root. CONJ. AF. Mod. and Con. Pers. Affix. Tense. Early. JLate Temp. Af. Vow. S. 1. mon (ea)* eo _ _ _ moneo. aj 2. mon (ea) e — S mones. o-^ S.- mon (ea) e — — t monet. ^1 P.I. mon (ea) e — — mus monemus. 2. mon (ea) e — — tis monetis. 3. mon (ea) e — — nt monent. S. 1. mon (ea) e ba m monebam. ^ 6 2 mon (ea) e ba — s monebas. i-B 3. mon (ea) e ba — t monebat. l|. P.l. mon (ea) e ba mus monebamus. 2. mon (ea) e ba _ tis monebatis. 3. mon (ea) ' ba — nt monebant. S. 1. mon (ea) e bo monebo. aj 2. mon (ea) e bi — s monebis. £•5 3. mon (ea) e bi t monebit. 3 '^ P.l. mon (ea) e bi — mus monebimus. ^1 2. mon (ea) e bi tis # monebitis. 3. mon (ea) e bu — nt monebunt. SI. mon (ea) ea a _ m moneam. ^^ 2. mon (ea) ea a — s moneas. 1i 3. mon (ea) ea a — t moneat. ^ a P.l. mon (ea) ea a — mus moneamus. SJ 3 pis' 2. mon (ea) ea a — tis moneatis. 3. mon (ea) ea a — nt moneant. S.l. mon (ea) e re _ m monerem. -» > 2. mon (ea) e re — s moneres. 11 3. mon (ea) e re — t moneret. 11 P.l. mon (ea) e re — mus moneremus. If 2 mon (ea) e re _ tis moneretis. 3. mon (ea) e re nt monerent. S.2. mon (ea) e mone. « 2. mon (ea) e — — to moneto. .& 3. mon (ea) e to moneto. i P. 2. mon (ea) e — _ te monete. & 2. mon (ea) e — — tote monetote. s 8. mon (ea) e — — nto monento. * In this conjugation the conjugational affix is carried back no farther than to ea, the form resulting from aja by the coayersion of j into i, and its combination with initial a to form e. 44 ANALYSIS OP THE VERB. ANALYTIC. SYNTHETIC. Mood and Tense. Root. CONJ. AF. Mod. and Temp. Af. Con. Vow. Case Affix. Early. Late. Pres. Inf. Pres. Part, (ierund. Fut. Part. (Pas.) mon mon mon mon (ea) (ea) (ea) (ea) e e e e n(t) nd ndu — e(Dat.) S (Nom.) i(Gen.) S(Nom.) monere. monens. monendi. monendus. TABLE IV. Conjugation of Audio. Active Voice. Present System. analytic. SYNTHETIC. Mood and Pers. Root. CONJ At. Mod. and Tense. Early. Late. Temp. Af. S. 1. aud (ia)* io audio. „ OJ 2. and (ia) i — s audis. 3. aud (ia) i — t audit. m c4 P.l. aud (ia) 1 — mus audimus. /^' "S 2. aud alld (ia) i tis auditis. 3. (ia) lU — nt audiunt. S.l. aud (ia) ie ba m audiebam. "Sg 2. aud (ia) ie ba s audiebas. 'cl 3. aud (ia) le ba t audiebat. 1" P.l. aud (ia) ie ba mus audiebamus. si 2. aud (ia) ie ba tis audiebatis. •n^ 3. aud (ia) ie ba nt audiebant. S.l. aud (ia) ia a m audiam. „, ^ 2. aud (ia) ia i(a) s audies. 2.E: 3. aud (ia) la i(a) t audiet. P.l. aud (ia) ia i(a) mus audiemus. Cx -o 2. aud (ia) ia i(a) tis audietis. 5 3. aud (ia) ia i(a) nt audient. S.l. aud (ia) ia a m audiam. -^ 2. aud (ia) ia a s audias. II 3. aud (ia) ia a t audiat. P.l. aud (ia) ia a mus audiamus. 2. aud (ia) ia a tis audiatis. 1 3. aud (ia) ia a nt audiant. * In this conjugation the conjugational affix is carried back to la, the form resulting from aja by the conversion of the initial vowel into i, and its union with j (also con- verted to i), to form i long. ANALYSIS OF THE VERB. 45 ANALYTIC. SYNTHETIC. Mood and CONJ. Af. Mod. and Pers. Affix. Tense. Early, j Late. Temp. Af. S. 1. aud (ia) i re m audirem. -s^ 2. aud (la) 1 re s audires. ^1 3. aud (ia) 1 re t audiret. P.l. aud (ia) 1 re mus audiremus. It 2. aud (la) 1 re tis audiretis. 3. aud (la) 1 re nt audirent. S.2. aud (ia) i _ audi. g 2. aud (ia) 1 — to audito. •-5 3. aud (ia) i — to audito. ^ P. 2. aud da) i — te audite. c, 2. aud (ia) i _ tote auditote. a 3. aud (ia) iu — nto audiunto. Mood an i Tense. Case Affix. Pres. Inf aud (ia) i r e (Dat.) audire. Pres. Par t. aud (ia) ie n(t) S (Nom.) audiens. Gerund. aud (ia) ie nd i (Gen.) audiendi. Fut. Pari . (Pas.) aud (ia) ie ndu S (;Nom.) audiendus. TABLE V. COXJUGATIOX OF Rego. Active Voice, Present System. ANALYTIC. SYNTHETIC, Mood and Pers. Root. CONJ Af. Mod. and Pers. Affix. Tense. Early, Late. Temp. Af. S. 1. reg (a) o rego. ->^ S 2. reg (a) 1 — S regis. c-r .,?• reg (a) 1 — t regit. is P.l. reg (a) i — mus regimus. £■■2 2. reg (a) i — tis regitis. 3. reg (a) u — nt regunt. S.l. reg (a) e ba m regebam. o £ 2. reg (a) e ba s regebas. 1-^ 3. reg (a) e ba t regebat. 11 P.l. reg (a) e ba mus regebamus. 2. reg (a) e ba tis regebatis. H-l 3. reg (a) e ba nt regebant. S.l. reg (a) a a m regam.