(Aen. VII-XII) in 1862; and vol. iv (Prolegomena) in 1866. The faults of this great edition (well criticised by Professor Conington in an essay now appended to his 3rd volume) arise mainly from undue suspicion of the received text, and undue confidence in critical sagacity to correct, transpose, and emend: but its solid merits far outweigh defects, and make it an important epoch in the history of the text, the basis (so far as can be foreseen) of subsequent criticism. III. ORTHOGRAPHY 1. The increased knowledge of MSS. and inscriptions which the present century (one might almost say, the present generation) has brought forth makes it impossible for any editor of a Latin classic to profess either ignorance or disregard of disputed questions in Latin orthography, or to acquiesce in the established spelling which satisfied our grandfathers. Barbarisms such as 'coelum,' 'coena,' 'lacryma,' 'sylva,' 'moereo,' introduced by the Italian scholars of the Renaissance 'to derive those words preposterously from the Greek,' are by this time almost unanimously discredited: but there are many other words, on the spelling of which the evidence of MSS. and inscriptions is practically conclusive, which continue to greet us in their wrong dress. The following is a representative, though not complete, list of such words : I. Words as to which MSS. and Inscriptions are unanimous, or nearly so (a) 'Querella,' 'loquella,' 'sollemnis,' 'Iuppiter,' 'littera,' 'quattuor,' 'bracchium,' damma,' 'ammentum'; 'baca,' 'belua,' 'conecto,' 'coniti,' 'conubium,' 'litus,' 'milia,' 'paulum,' 'religio,' 'reliquiae.' (b) 'Condicio,' 'dicio,' 'solacium,' 'ocius'; 'nuntius,' 'setius,' ' autumnus.' (c) 'Umerus,' 'umor,' 'umidus'; 'erus,' 'erilis'; 'harena,' 'harundo,' 'haruspex,' 'hedera' (also 'edera '?), 'Hiberus,' 'Hister,' 'aënus' ('ahenus'), 'incohare.' (d) 'Adicere,' 'conicere,' 'deicere,' and so all compounds of 'iacio,' except 'disiicere' (? 'dissicere'), [some good MSS. however : For fuller information on this subject see Munro's Lucretius,' Notes I, Introduction Ramsay's 'Mostellaria' of Plautus, Prolegomena II, pp. xvi-xlix; Kennedy's 'Virgil,' Appendix D, pp. 628-630 (second edition); Public Schools' Latin Grammar, Appendix A. give 'proiicit,' 'deiicit' or 'deiecit' (see on Ecl. iii. 96), and it is not clear that 'deicit,' etc. were universal in Virgil's time]. (e) 'n' before 'd' in 'quendam,' ' eundem,' etc.; but 'circumdo,' ‘iamdudum'; 'm' before 'q'in 'quemquam,' ‘tamquam,’‘umquam': 'nequiquam.' (ƒ) ‘Caelum,' 'caenum,' 'caespes,' 'faenum,' 'foedus,' 'frena,’ 'glaeba,'' scaena' (' scena '), ' proelium,' ' saepes': 'cena,' ' ceteri,' 'femina,' 'fenus,' 'lēvis,' 'obscenus,' ' silva,' 'Vergilius.' (g) 'Hiemps' (phonetic insertion of 'p' sound, as in 'sum-p-si,' 'tempto.' Thus far our course is, or should be, clear. There are, however, many words and forms in the spelling of which, as Mr. Munro says, 'variety was the rule of the ancients'; whose orthography, that is, gradually changed during the period from Ennius to the first century of the Empire, so that inscriptions of that time, and MSS. of authors that fall within it, present us with varying forms of the same word. Such fluctuations, it is probable, were stil going on when Virgil wrote, though Cicero had done much to establish a uniform literary standard of spelling and it is not strange that Virgilian MSS. should present us with varying forms. Whether Virgil himself, on a final revision of his own work, would have allowed so great a variety as meets us in modern editions of his text, is another question, the answer to which can only be matter of opinion, never of demonstration. The following list shows the principal heads of varying orthography : II. Cases in which ancient usage seems to have varied, especially in the Augustan age, when changes in literary spelling were in progress: (a) Assimilation of prepositions-e. g. 'inperium' or 'imperium,' the former representing etymology, the latter pronunciation. Words in commoner use adopted the latter; thus MSS. of Virgil always have 'imperium,' but 'impius' and 'inpius,' 'immortalis and 'inmortalis,' 'compleo' and 'conpleo' are found; ' exsto' and ́exto,' etc.; and so in compounds of 'ab,' 'ob,' 'sub,' 'ad.' (6) Accus. plur. in '-es' or '-is'; the former being more common in words of very general use, e. g. 'ignes,' 'vires,' 'aures.' [Inscriptions also give ' -eis.'] (c) 'o' or 'u' after 'v' (i. e. consonantal 'u'). The older Latins found difficulty in pronouncing 'uu' ('vu'), except where both sounds were vowel, as in 'tuus,' suus' and the older 'o' long kept its place after 'qu,' or 'u' ('v'), the former being some times changed to 'c.' Thus 'quom' or 'cum,' not 'quum'; 'secuntur' or 'sequontur'; 'equos' or 'ecus'; 'volvont,' 'novos,' 'alvos,' etc.; are correct in Plautus and Lucretius. But in Virgil's time the feeling against 'quu,' 'vu' was subsiding, and they were coming into general use. (d) 'u' or 'i' where the sound was intermediate between the two, and the spelling therefore uncertain; e. g. 'maxumus' or maximus,' ' lacruma' or 'lacrima,' 'lubet' or 'libet,' 'inclutus' or 'inclitus' (such forms as 'lacryma,' 'inclytus' are wrong); 'o' or 'e' in 'vorto,' 'verto,' 'vortex,' 'vertex,' etc.: 'i' or 'e' in 'protinus' or 'protenus,' etc. Ribbeck also writes 'haut,' 'set,' 'aput' as well as the ordinary forms in '-d': 'ps' as well as 'bs' in forms like 'obstipui,' 'obsto,' etc.; and always 'supter': 'quoi' as well as 'cui': and, where 'est' follows a vowel or 'm,' ' necessest,' 'ventumst,' etc. as usual in Plautus. But though all these forms had existed (see Ramsay's Introd. to Plautus, 'Mostellaria'), it is not clear that they were employed in Virgil's time. In dealing with such cases more than one course is open to us. (1) We may ignore scientific orthography and the evidence of inscriptions, and adopt the 'conventional' spelling of the Renaissance scholars, as it appears in ordinary grammars and text-books, leaving all questions connected with the history of changes in orthography for separate and more advanced study. This method, however, would commit us to so many positive and ascertainable errors (as indicated above), that it would in the present state of scholarship be simply retrograde. (2) We may profess, from the evidence of MSS. and inscriptions, to reproduce the orthography of Virgil's own time, and present the poems as nearly as possible in the shape in which they came from his hand. But can we do this? Inscriptions, no doubt, supply contemporary evidence: but in an age of transition in orthography, considerable deduction must be made for the ignorance or carelessness of the persons employed to cut them. The oldest existing MSS. were written 350 years after Virgil's time, and it is obvious that their testimony to the actual character of his own MS. can be little more than inference, more or less weakened by the chance of variety and error in successive generations of copyists. And the authority of grammarians, though it may date back to a time much nearer Virgil's own, is but of little value. They had all the dogmatism and love of theory of the most dogmatic modern scholar, with less than his opportunities of critical and philological know ledge their statements upon matters of orthography are often very unscientific and the text of their writings, moreover, is usually more corrupt than that of the great classics. Turn which way we will, it seems useless to pretend to reproduce the text exactly as Virgil wrote it. (3) A sounder method is that inaugurated by Lachmann on Lucretius, and since followed by Ritschl on Plautus, Ribbeck on Virgil, and a host of German scholars-viz. to sift as carefully as possible the existing MS. evidence, and from it make the best approximation to what the author might have written, taking as a rule the oldest form for which evidence, direct or indirect, is forthcoming. For a purely critical edition, such as Ribbeck's, whose aim is not commentary but textual criticism, this is the only reliable method: and the system adopted by Wagner in cases of varying orthography-viz. to decide on a priori principles what Virgil must have written-is justly condemned by Mr. Munro, in his Introduction to Part I. of his notes to Lucretius, as involving him in 'a sea of conjectural uncertainty.' (4) It seems, therefore, that the most obvious and the safest course for editors who do not profess a first-hand study of MSS. is, to adopt the latest results of critical investigation, and to reprint e. g. Ribbeck's text of Virgil, as Conington (in matters of orthography) does that of Wagner, or as an editor of Lucretius (were there now any room for him) would reproduce that of Lachmann or Munro. But Ribbeck's work, invaluable for its careful and patient collection of materials for textual criticism, is disfigured by much arbitrary transposition of lines, much unwarrantable suspicion of genuineness, and undue licence in the tempting but slippery field of conjectural emendation, accompanied sometimes by a curious insensibility to poetical feeling or even to common sense 1. His text must be the basis for any fresh edition; but to reproduce it as it stands seems impossible. A further question then occurs: Is it necessary, in dealing with the writings of a Latin classic for educational purposes, to maintain all the variety of spelling which MSS. undoubtedly exhibit—a 6 1 Illustrations of this may be found in Professor Conington's remarks on Ribbeck's Prolegomena appended to vol. iii. of his edition of Virgil. A typical example is Aen. xii. 55, where, in the description of Amata, selfdoomed to death, and clasping Turnus for the last time, ardentem generum moritura tenebat,' Ribbeck, in defiance of MS. authority, reads monitura, 'with intent to advise,' a singularly inappropriate word for the impassioned appeal which follows. variety so great that in an English classic we could not, and should It seems best to answer these questions in the negative with regard to most of the cases of varying orthography that fall under the heads above given in List II (p. xlv); and to write throughout 'impius,' 'compleo,' 'immortalis,' 'irrigo,' 'colligo,' etc.; '-es' in nom., acc. plur. of adjectives and substantives of 'i' and cons. stems, and participles; 'equus,' 'linquunt,' ' vulnus,' 'vulgo,' etc.; 'maximus,' 'verto,' etc. The two heads which embrace the greatest number of departures from the orthography now familiar in Conington's text are those of 'vu' ('uu') instead of 'vo-,' and '-es' instead of '-is,' in nom., acc. plur.: and in both these cases an examination of the MS. testimony collected by Ribbeck goes further than might have been expected to justify the course taken. Vu as well as vo is found in the best uncial MSS.; e. g. divum, vultu, vulgo, vulneribus. And with regard to -is or -es in nom., acc. plur., though the preponderance of MS. testimony is generally considered to be in favour of '-is,' the form -es is found in many words, and in some of the more common words (as 'cives,' 'puppes,' 'vires,' 'fruges') in a majority of cases where they occur. This form has thus a firm 'locus standi' in the text and we may ask whether, considering the gain of distinguishing between gen. sing. and nom., acc. plur., and the fact that in the period immediately succeeding Virgil's time such distinction became established, we are not justified in giving his text the benefit of it, and removing the (at first sight) unintelligible variety which that text exhibits in the 1 The form 'olle,' deliberately introduced as an archaism side by side with the later 'ille,' does not affect the point. T |