Qui rore puro Castaliae lavit

Crinis solutos, qui Lyciae tenet

Dumeta natalemque silvam

Delius et Patareus Apollo, Hor. Od. iii. 4. 61.

Horace has thirty-seven Odes in this Metre.

NOTE.

The metres of the Comic poets, Plautus and Terence, are too large a subject to be treated in this grammar. We will merely observe that many final syllables (ar, or, at, et, it), short in later poets, are lengthened by Plautus and Terence; Iambic words, on the other hand (such as habent, bonis), are often scanned as Pyrrhichs; the law of position is often violated; and long initial syllables slurred into short quantity, when they follow monosyllables or elided Pyrrhichs. Add to these licenses the most extensive synizesis and the free use of Spondee, Dactyl, Anapaest, even Proceleusmatic for Iambus or Trochee (always excepting the final foot), and it will be seen at once in how wide a field of rhythm the old scenic poets ranged. See p. 56.

APPENDIX.

A. LATIN ORTHOGRAPHY.

ORTHOGRAPHY is defined by Suetonius (Oct. 88), as the form and method of writing taught by grammarians,' and by Quintilian (i. 7), more shortly, as 'the science of writing correctly.' We may blend the two definitions, and say that Orthography is the science of writing in correct form the words of any language.'

The Greeks settled their orthography with reference to four considerations: (1) analogy; (2) etymology; (3) dialect; (4) history. The Latins left dialect out of question, but had regard to the other three points. Yet, owing to the fluctuating character of their language, and its many changes during the seven centuries between the first Punic war and the fall of the Roman empire, the settlement of a solid Latin orthography is a work of difficulty. The labours of Ritschl, Lachmann, and others have indeed during the last few years thrown much light on this subject. Ritschl justly selects the age of Quintilian's great work (about the close of the first century, A.D.) as the standard of Latin orthography; but, unfortunately, it is only by an inductive process, often uncertain, that the forms of words can, generally speaking, be referred to this age. Inscriptions, of course, have the greatest value; but they are often inconsistent even when contemporaneous. The earliest MSS. are several centuries later than the Christian era; and they also disagree. The opinions of old grammarians are not less various. Hence it often happens that the classical form of a word can be determined only by a balance of conflicting evidence; and different minds will strike the balance differently. A few probable results, compendiously stated, must suffice here. The student may compare Munro's Introd. to Lucretius, and Ribbeck's Proleg. to Virgil. Here forms assumed to be most classical are named first: others of nearly equal authority are added with 'or:' those of minor authority are within brackets; those which seem inadmissible follow 'not,' and are in italic type.

1) a, e defatigo or defetigo; depeciscor (depaciscor).

2) e, ae, oe: caecus, not coecus; caelum, heaven, not coelum; also caelum &c. graving-tool; caementum (cementum); caerimonia, not ceremonia; caespes, not cespes; Camena, not Camoena; cena &c., not coena &c.; ceteri, not caeteri; faenum (fenum ?), not foenum; fecundus, not foecundus; femina, not foemina; fetus &c., not foetus; fenus (foenus); foedus, treaty; glaeba; heres, not haeres; lēvis, not laevis; maerere and maestus, not moerere and moestus; oboedire (obedire); obscenus, not obscaenus obscoenus; Paelignus,

not Pelignus; paelex, not pellex; paenitet, not poenitet, but poena ; paenula, not penula; pomaerium (pomerium); prelum, not praclum; proelium, not praelium; raeda, not reda rheda; saeculum, not seculum; saepes &c., not sepes &c.; scaena (scena).

3) e, 1: benevolus (benivolus); deminuere &c., not diminuere &c.; di (dei), dis (deis); genetrix, not genitrix; heri (here); intellego (intelligo); neglego (negligo); protinus or protenus, but quatenus; valetudo (valitudo); Vergilius, not Virgilius.

As respects is (eis) or -es, Accus. Plur. of I-nouns, admitting that in the republican age -is was the more usual, as it is certainly truer in formation, yet we believe that, before the age of Quintilian, -ēs was in general use; and this, with its superior convenience, has led to its frequent retention. On -ě or -i in Abl. of I-nouns, see § 24. 5.

4) i, u: The middle tone between ĭ and ŭ (see § 11. p. 8; § 12. p. 33), led to the existence of a large number of double forms: aestimare (aestumare); Brundisium (Brundŭsium); inclutus (inclitus); lŭbet &c. or libet &c. ; recuperare (reciperare); Dat. Pl. of Decl. 4: grad-ibus &c. (gradŭbus &c. § 25); maximus (maxŭmus), and all Superlatives; vicensĭmus (vicensumus), and other Ordinals in -ĭmus (-ŭmus); so maritimus (maritumus); monimentum or monumentum ; tegimen or tegumen; and other similar derivatives. But the ú-form in many of these was archaic in the Augustan and following age, which wrote carnifex rather than carnŭfex; lacrima rather than lacruma; clipeus rather than clupeus; optimus rather than optimus, except perhaps in old formulas; mancipium, not mancupium; victima, not victuma.

5) e, u: -endus (-undus) in Gerundive forms: -undus was the ancient form, but superseded by -endus in the Imperial age. 6) o, u adulescens (Noun), adolescens (Part.); epistula or epistola; suboles (soboles). The earlier Latins, even to the Augustan age, wrote o rather than u when u preceded: but u was received under the emperors: hence vult (volt); avus (avos); equus (equos), &c. But Ribbeck in Virgil almost always avoids uu, vu.

See 12.

The

7) e, o: vertere (vortere); versus (vorsus); vertex (vortex). forms in o are comparatively archaic. But fenoris or feneris, feneror; iecinoris or iecineris; pignoris or pigneris, pigneror.

8) 1, yy is not properly a Latin letter, but introduced in Cicero's age to represent Greek v. Therefore, such forms as clypeus, hyems, inclytus, ocyus, satyra, stylus, sylva, Sylla, are now justly exploded, the true forms being clipeus (clupeus), hiemps, inclutus (inclitus), ocius, satira, stilus, silva, Sulla. But, where Greek v is represented byy, this letter holds its proper place: lyra, Nympha, Syrus, Syria, Tyrus, &c.

9) guo or go: lingĕre (linguere); ningit (ninguit); stinguere and compounds, not stingere; tingere (tinguere); ungère (unguere), but unguentum, unguen; urgêre (urguēre).

10) g or c vicensimus (vigensimus), trigensimus or tricensimus, but quadragensimus, &c.; so ducenti, trecenti, sescenti, but quadringenti or quadrigenti; quingenti, &c. ; viceni, triceni, but quadrageni, &c.; duceni, treceni, sexceni or sesceni, but quadringeni, quingeni, &c. See NUMERALIA, § 33. Cycnus or cygnus; Cnosus

or Gnosus.

11) gn or n: nasci not gnasci; natus (gnatus), but agnatus, cognatus, &c. ; noscere not gnoscere; but agnoscere, cognoscere, &c.

12) c or q (qu): cotidie or cottidie, not quotidie; coquus (anc. cocus, coqus); equus (anc. ecus, equs, equos); pecunia (anc. pequnia); locutus (anc. loqutus); secutus (anc. sequtus); loquuntur, sequuntur (anc. locuntur, secuntur); cui (anc. quoi, quoei); cur (anc. qur, quor); quum or cum, conj. (anc. qum, quom); cum, prep. (anc. qum, quom). The form cum is good for preposition and conjunction ; quom was used for both to the Augustan age: after which the dislike of uu seems to have gone out of fashion; and the form quum is Ribbeck often used for the conjunction. But qu was uttered as c.

in Virgil commonly edits ecus, ecum (or quos, quom) locuntur, secuntur, &c., instead of the forms with qu.

13) b for v: ferbui or fervi; to avoid vu.

14) ʼn initial present or absent: Hadria, not Adria; alucinari (halucinari); Hammon (Ammon); harena (arena); harundo or arundo; haruspex or aruspex; hariolus, not ariolus; have (ave); hedera, not edera; erus, era, or herus, hera; heres (eres); herciscere, not erciscere; holus (olus); Hiber (Iber); Hister (Ister); umēre, umor, &c., rather than humere, humor, &c.; umerus, not humerus.

15) h interior: aeneus, &c. (aheneus, &c.); cohors or cors; incohare (inchoare); nihil or nil; prendo (prehendo); vemens, not vehemens. 16) euphonic p: compsi, comptum, &c. (comsi, comtum, &c.), and others; hiemps (hiems).

17) ci or ti before a vowel. Authority favours dicio, condicio, solacium, patricius, tribunicius, &c., not ditio, &c.; and contio, fetialis, inAlso condutiae, nuntius, nuntiare, &c., setius; not concio, &c. vitium rather than convicium; suspitio (Subst.), rather than suspicio; but the forms of these with ci were also used. See Corssen, I. 56. 18) b or p: caelebs (caeleps): urbs (urps), &c., obsonium, obsonari (opsonium, opsonari); obtulit (optulit); subter (supter). But bs was sounded as ps, bt as pt.

19) d or t final. In Quintilian's time the endings in d, haud, sed, apud, &c., had become general in preference to the archaic t, haut (hau), But d final was sounded as t. set, aput, &c.

20) -icere or -iicere. The compounds of iacere have been elsewhere noticed. To the Augustan age the single 1 seems to have prevailed: But in imperial times ii adicere, deicere, conicere, reicere, &c. was at least admissible: adiicere, coiicere or coniicere, &c., the former 1 being a consonant. And in adicere, &c. i did double duty as consonant and vowel = ad-yi-cere, &c.

21) ct or t: artus not arctus; autumnus, not auctumnus; but auctor. 22) n kept or omitted: conectere, conexus, coniti, conivēre, conixus,

conubium, not connectere, &c. (M. Lucr. i. 633): coniunx (coiunx coiux): n before s was liable to elimination: thus Numeral Adverbs in -ens passed into -es; the earlier form being, however, generally preserved: quotiens (quoties); totiens (toties); miliens (milies), &c. So vicensimus (vicesimus). Such omissions of n abound in Inscr. See Corssen, I. 249, &c. as cosol for consol, cesor for censor. 23) Consonants singled or doubled on this point documents are especially discrepant. We find Britannia or Brittannia; cottidie or cotidie; Iuppiter (Iupiter); littera (litera); loquella or loquela; querella or querela: but medēla, suadēla, &c., also cautela, tutēla, reliquiae (relliquiae), but relicuus (later reliquus); religio (relligio): causa (caussa): paulum (paullum). But nummus (nūmus); bracchium, not brachium; Messalla, not Messala; sollemnis, not solennis;

sollers, &c., not selers, &c. ; sollicitus, not solicitus, &c. Yet anulus; rather than annulus; culeus rather than culleus, &c. ; litus, not littus; ilico rather than illico; vilicus rather than villicus; stilicidium. Pliny wrote mille but milia; Augustus, however, writes millia, milliens; and Corssen, I. 226, prefers the 11. Reppuli,

repperi, rettuli, seem better than repuli, &c., which are, however, used.

24) Assimilation: quidquid (quicquid); quicquam (quidquam), quicque (quidque); quamquam (quanquam); tamquam (tanquam); umquam, numquam (unquam, nunquam); -cumque (cunque); quendam not quemdam; eundem not eumdem; tantundem not tantumdem; quorundam not quorumdam; eorundem not corumdem. When m remains before the guttural or dental, it is sounded as n : quamquam quanquam when uttered.

25) Assimilation of Compounded Prepositions.

This is received or

rejected in MSS. and Inscr. with such apparently free option in most instances, that no safe rules on the point can be laid down. Thus we read in imperial times adlectus and allectus, collapsus and conlapsus, collegium and conlegium, illustris and inlustris, impendium and inpendium, irritus and inritus, even imperium and inperium, &c. On the whole assimilation prevails. See Brambach, Neugestaltung der Latein. Orthographie, p. 300, &c.

26) Not less uncertainty exists in regard to words beginning with s, when compounded with ex (ecs: c. èk, §). We find exspectare and expectare, exspirare and expirare; exsequi, exsequiae, and exequi, exequiae; exsecrari, &c., and execrari, &c. ; exserere and exerere; exsilire and exilire; exsultare and exultare; especially exsul, exsulare, exsilium with exul, exulare, exilium. In some words, it is perhaps better to retain the s; but exul, &c., have ample authority; and excidium (from exscindo) is better established than exscidium.

27) Accessory Note:

'We have now (writes Mr. Munro) an accurate transcription of the large fragments of the Ancyra monument, containing no doubt an exact copy of the "Res gestae " of Augustus, which, as Tacitus (Ann. i. 11) tells us, he had written out with his own hand, a short time probably before his death. The spelling is interesting, as Suetonius says that Augustus was a purist on such points.

'He always admits uu: rivus, vivus, as well as annuus; but he writes Phrates, praerant.

'In Gen. Plur. we find denarium, sestertium, deum, nummum ; triumvirum, but xv virorum. In Gen. Sing. always 1, not li; proeli, Iuli, Pompei, congiari. In Dat. Abl. Plur. of Decl. I and 2, both is and iis: dis, colonis, provincis, &c. ; but also consiliis, &c.; both municipis and municipiis. Dalmateis, emeriteis, quadrigeis: but oftener is in Dat. and Abl. Plur. Sometimes is, but oftener es in Accus. Plur. of 3rd Decl.: once pluris in Nom. Plur.

'Honos, incohare, Messalla, plebis Gen. Sing., but plebei Dat, sescenti, valetudo.

Always 1, not u, in the fluctuating instances: legitimus, septimus, frequentissimus, reciperare, manibiae, &c. Compare what Sueto nius (ch. 87) says of his writing simus for sumus. Perhaps it was this love of consistency which makes him always spell millia,