m/*'W.VA*'w«»i«*^fgpp»^KwsMa^ iaiilpi^ iNiMjiii UC-NRLF *B 2Tfl o^e r: ii; mU'.'' VEE8IFICATION, LONDON : PRINTED BY SPOTTISWOODE AND CO., NKW-STRKKT SQUAEB AND PARLIAMENT STREET A TREATISE ON VERSIFICATION ■ BY GILBEET CONWAY. Qui studet optatam cursu contingere metam, Multa tulit fecitque puer; qui Pythia cantat Tibicen, didicit prius, extimuitque magistrum.' Hoi!. LONDON : LONGMANS, GREEN, AND CO. 1878. {All rights reserved.) 9 0^ C167 PEE F AC E The following treatise rests on. the assumption that every artist should begin by mastering the rudiments of his art : and the author's main object is to assist those who, as- piring to become English poets, may find a difiiculty in deciding what are the laws which rule, or ought to rule, the mechanism of English verse. 386 THE SUBJECT STATED. *Si quis autem, ex nostratibus prsesertim, hoc totum quicquid est operis penitus omitti posse credat, neque rem magnam esse dicat vernaculam linguam callere, ej usque minutias observare, hoc solum regero ; — multa esse quae, quamvis cognita, non magnam mereantur laudem, eadem taraen ignorata non leve possunt dedecus imprimere.' (Joannis WaUisii, S.T.D., ' Grammatica Linguae Anglicanse.' Auctoris ad Edit. v. praefatio. Mnr.TriY \ Errata. P. 51, note *, line 28, for syllable read vowel. „ 81, line 1, for seem'd to eye read seera'd to the eye. Treatise on Versification. not easily impngueu ; uul nc ^^j „v... ...^ . sense : for as poetry comprises three elements, thought, language, and verse, it would be going too far to pronounce worthless all poetical composition which fails to attain supreme excellence in each of these. Fine thoughts, well expressed, will always have an intrinsic value of their own, even though happening to be embodied in ill-shapen verse ; and all we are en- titled to say, when this happens, is, that they lack the charm which verse is capable of imparting. Some people, indeed, say that they are indiiFerent to defects of ver- sification, as ' they read poetry only for the thoughts ; ' but this is tantamount "^ to an acknowledgment that they have no perception for poetry ; for it is | verse which makes the difference between poetry and prose ; and if the ! thoughts only are to be considered, there is no need of verse at all : the / thoughts would do just as well in prose ; nay, far better in good prose than in bad verse. Others again say ' that they prize the jewel, but do not care for the setting : ' but the analogy here implied is a false one ; for the setting of a jewel is an accidental and external circvimstance, whereas verse is an element of poetry not less essential and intrinsic than the thought itself. The simple truth is, that even among well educated people, there are many who have no ear for poetical numbers, just as there are many, of the same class, who have none for harmony, or tune, in music. But what B THE SUBJECT STATED. ' Si quis autem, ex nostratibus prsesertim, hoc totum quicquid est operis penitus omitti posse credat, neque rem magnam esse dicat vernaculam linguam callere, ej usque minutias observare, hoc solum regero ; — multa esse quae, quamvis cognita, non magnam mereantur laudem, eadem taraen ignorata non leve possunt dedecus imprimere.' (Joannis WaUisii, S.T.D., * Grammatica Linguaj Anglicanse.' Auctoris ad Edit. v. prsefatio. MDCXCIX.) These pages treat of Versification ; a subject which, if little studied or appreciated among us, is one, however, which con- cerns all poets not a little ; because though in most things there is some medium between good and bad, there is none between them in verse.* But an English student who takes * When Horace says that if a poem is not quite up to the highest mark, it lapses to the opposite extreme, — ' Si paulum a sum mo discessit vergit ad imum,' — he seems here to be regarding poetry with special reference to its distinctive characteristic, verse ; and his words, taken in this sense, express a judgment not easily impugned ; but we may well hesitate to accept tbem in a wider sense : for as poetry comprises three elements, thought, language, and verse, it would be going too far to pronounce worthless all poetical composition which fails to attain supreme excellence in each of these. Fine thoughts, well expressed, will always have an intrinsic value of their own, even though happening to be embodied in ill-shapen verse ; and all we are en- titled to say, when this happens, is, that they lack the charm which verse is capable of imparting. Some people, indeed, say that they are indifferent to defects of ver- sification, as ' they read poetry only for the thoughts ; ' but this is tantamount to an acknowledgment that they have no perception for poetry ; for it is verse which makes the difference between poetry and prose ; and if the thoughts only are to be considered, there is no need of verse at all : the thoughts would do just as well in prose ; nay, far better in good prose than in bad verse. Others again say ' that they prize the jewel, but do not care for the setting : ' but the analogy here implied is a false one ; for the setting of a jewel is an accidental and external circumstance, whereas verse is an element of poetry not less essential and intrinsic than the thought itself. The simple truth is, that even among well educated people, there are many who have no ear for poetical numbers, just as there are many, of the same class, who have none for harmony, or tune, in music. But what B Z DEFINITION OF TEEMS. up this study will get from it more harm than benefit if he looks no further than his native language : verse cannot be good if it be not, primarily, found correct ; and nothing can be so found save by reference to some standard of correct- ness, based on acknowledged laws : now, we have no standard of our own whereby to test English verse, seeing that, from first to last, our poets have acknowledged, as binding on them, scarce any law whatever. In the absence, then, of any consistent standard of versification, to be drawn from the practice of our poets, it may be worth while to enquire whether there be such in any other language ; and if such in some other there be, then whether it be applicable to ours ; and if so, then lastly, how far the system used by us accords therewith. But before going further, we need to fix the basis of oui* enquiry ; and the fii'st thing to be done, with that view, is, to define the sense of certain terms.* By Verse, I mean words arranged on some definite plan tending to melodious effect of a distinctive kind ; By Metre, the various distinctive forms which the melody j of verse takes ; By the Laws of Verse, the order of things on which the melodious effect depends ; By Faultiness, a state of things which results from in- attention to the laws of verse ; By Correctness, the quality of being free from fault ; then ? Because some are unable to distinguish between right and wrong in art, should art therefore abandon itself to anarchy ? * Non quivis vidit immodulata poemate judex : Idcircone vager, scribamque libenter ? ' * With reference to Aristotle's ' Treatise on Poetry,' an Italian poet, who had made a special study of it, thus writes : * Converebbe qui, per I'intelligenza successiva del testo, determinarsi su le proprie significazioni delle parole "Metro," " Ritmo," "Armonia," "Melodia," e " Modi ; " ma gl'in- terpreti son cosi mal concordi fra loro, e gli antichi scrittori, ed Aristotele medesimo, se ne vaglion cosi promiscuamente, che diventa difficillissima impresa I'evitarne la confusione. Pure io, senza spacciare, per sicura, la mia eentenza, confesserb ingenuamente in qual senso, spiegandole, mi sia paruto di urtar meno in manifeste contradizioni.' (Metastasio. Estratto della Poetica d' Aristotele, L. 1.) t Musical terms apply but in a partial sense to verse ; and there is no need to define the difference between the melody tff verse and that of music proper. QUANTITY AND ACCENT. 3 By Harmony, (as applied to verse), those effects of sound which, arising from * a just adaptation and proportion of parts to each other,' * are satisfactory to the ear ; By Prosody, the laws which regulate the measure of sound, or the force of sound, in syllables ; By Quantity, the measure of sound due to syllables ac- cording to their relative value ; By Relative Value of Syllables, proportion of one to two, and two to one ; (one long syllable being the equivalent of two short, and two short of one long) ; By Accent, a stress necessarily laid on certain syllables ; f By Emphasis, a manner of pronouncing words, with a view to mark pointedly some distinction which they convey. J Thus, accent is an effect of prosody, emphasis of ex- pression; the one affecting syllables, the other words ; the one unavoidable, the other optional. There are two distinct principles, then, on which prosody may be based, quantity and accent ; (that is, syllables may be regarded either with reference to their greater or less length of sound, or to their greater or less force of sound) ; and the laws of verse will vary according as the prosody of a language is based on one or the other of these principles. The prosody of the Greek and Latin tongues is based on quantity ; § that of all modern tongues (except the French) * Dr. Johnson. t ' Dwelling on a syllable,' * a more forcible utterance,' a * greater intensity of voice,' * the incidence of the voice on a particular syllable,' * a smarter stroke of the voice,' ' intensio,' ' impressio,' ' percussio,' ' un frappement plus sensible,' ' une syllabe sur laqueUe on appuie plus fortement que sur les autres,' — these are all different ways of expressing the same thing, X The manner consists in an unusual enforcement of accent, and in change of note, perceptible, chiefly, in final syllables. The term ' final ' here includes monosyllables. § ' The Public School Latin Grammar ' (p. 5) defines quantity to be * the time of uttering a syllable ; ' and the unit of time being given, we see what is meant ; not so, however, when elsewhere, in the same work (p. 448), the term is applied to syllables of modem tongues ; for as quantitive syllables are' long' and 'short,' we need to be told in what sense ours are such. Nothing is ' long,' and nothing ' short,' save by reference to something else which is less ' long,' or less ' short ; ' and the ' something else ' has in this case to be defined. Unless, therefore, it can be said that English syllables have fixed proportions relatively to each other, all terms which imply they have should be excluded from our prosody, as tending to cause confusion of ideas. b2 4 WHAT ACCENT IS NOT. on accent : to those, then, one and the same code applies ; to these, one and the same, also ; but different from the other. The prosody of our language being based on accent, let us strive, at starting, to understand not only what accent is, but what it is not. I hold it to be stress, and nothing else ; but many writers contend that it combines with stress an elevation of the tone to a higher note in the musical scale. Among the chief of those who take this view are Ben Jonson,* Dr. Foster,f Dr. Gally,J Mr. Mitford, § Walker, || Professor Blackie,^ and Professor Newman ; ** nor are there wanting some foreign prosodists who seem to countenance this assumption as regards their own language, "ff * ' English Grammar,' cap. viii. His language on this point is, how- ever, not precise. t ' Treatise on Accent and Quantity',' and ' Replj' to Dr. Gaily.' X * Dissertation on Greek Accents.' Dr. Gaily seems to hold that accent is movement to a higher note, and that the movement causes stress ; others take the converse view ; namely, that stress causes the upward move- ment. ' Enquirj' into the Principles of Harmony in Language.' Observations on Greek and Latin Accent and Quantity.' ' Classical Musaeum,' vol. iii. •* ' Classical Musaum,' vol. i. p. 25. Speaking of the Greek language, Mathiae says (G. Gr., Blomfield's tr. p. 51) that the raising of the tone, in which accent consists, can no more lengthen a syllable than a quaver can become a crochet by being sharpened, or by being raised a semitone or a tone ; and Dr. Foster and James Harris (Philolog. Enq.) hold, in sub- stance, the same language, alike as to Greek and to Latin accent ; while both admit that English accent has the effect of long quantity : but if it be true, as Professor Newman says, that no essential difference can be estab- lished between Greek and English accent, then it must follow, either that the account given of Greek accent by Mathiae, Foster, and Harris, and of the Latin accent by the two last, is wrong, or that English accent is elevation of musical pitch, without any effect of long quantity. ft For example : ' Dans chaque mot compose' de plusieurs syllabes, il y en a toujours une,sur laquelle la voix, enpronon§ant le mot, se fait entendre plus fortement que sur les i.uties. Cette elevation de voix, ce frappement plus sensible sur une syllabe, qui ccnsiste en un coup de gosier qui eleve le ton d'un degre, pour retomber ensuite sur le ton dont ilest parti, est precise- ment ce qu'on appelle " accent tonique." En entendant un Italien prononcer le mot " sovrano " I'oreille s'apergoit que la voix s'eleve sur la syllabe " vra," ce qui fait entendre que, dans ce mot, I'accent tonique se trouve sur la pe'nultieme syllabe.' (See preface to Alberti's Dictionary. The same passage occurs also in Baggioli's Grammar). Again : ' Dans chaque mot compose de plusieurs syllabes, il y en a toujours une qu'on prononce avec plus deforce que les autres. Cette elevation de voix, rendue plus sensible sur une syllabe dans la prononciation d'un mot, s'appelle accent tonique L'accent, qui est I'ame de la pro- nonciation, est Velevation, plus ou mains forte, de la voix sur certaines syllabes.' (Zotti, Gr. Ital. pp. 274-5). AN EXPERIMENT SUGGESTED. 5 The controversy turns on a matter of fact ; and the fact may be tried by a very simple experiment. Let any one, then, place himself before the key-board of a pianoforte, and choose a note suitable to the usual pitch of his voice : let him then sound, quite naturally, to that note, any word of two or more syllables, and observe whether there be any change of pitch, as between accented and unaccented syllaA ^ -y bles : he will find, I feel sure, that each wQrd is pronounced ) ^ ^ in ^rict monotone : on the other hand, let him pronounce V words singly, or in sentences, with any elevation he pleases from a semitone to a fifth, on each accented syllable ; and then say what he thinks of the effect : he will find it neither speech nor song; and to him will apply the sarcasm of Caesar, — ' Si cantas male cantas, si le^is cantas.' * In fact, prosodians of modern languages almost invariably define accent to be what Prisciansays the Latin accent is, namely, ' Certa lex, vel r* gula, ad elevandam et deprimendam sjdlabam uniuscujusque particulae orationis.' (De Accentibus Liber), Thus, the question is embarrassed at its outset by the use of ambiguous terms ; for a doubt immediately arises as to what is meant by this ' elevation and depression ' of syllables. We may either mean that by the act of passing from one syllable to the next the voice ascends from a lower note to a higher in the musical scale ; or we may mean that, without change of degree, one syllable is uttered with more intensity of voice than another : for any sound may be intensified or relaxed, to whatever extent we please, on one and the same tone. Now, in the passages above quoted, it seems to be assumed, or implied, that enforcement of syllabic sound, and ascension of pitch, are the same things, or correlative ; and a similar fallacy runs through Dr. Foster's elaborate treatise on the ancient accent and quantity. Dr. Foster defines accent to be * intensio vocis ; ' a definition I accept : referring then (p. 80) to the Greek verb Tftvio, and its derivatives •Tcia-i,?, t6i/o?, tenores, tonic, tones,' he is at great pains to prove that these words, used prosodially, are restricted to signify extension in height only ; forgetting that they apply not more to the acute accent than to the grave, which is not by any one said to have the property of ascending. However, without caring to discuss with him theso points as they regard the ancient tongues, I am content to say that when Quintilian speaks of 'intensio vocis, remissio,' (intensity and relaxation of voice), he uses words which exactly suit the sense by me attributed to the word accent in modern tongues. * In Dionysius's ' Discourse on the Structure of Speech,' there is a passage ^ here deserving notice : — ^ AiakeKTOV (x-kv ovv fiekos evl {JLerpeiTaL SiacnjutaTt tc3 keyofieyco Sta irevre, cos ^ eyyiara' Koi ovre eTTireiveTat, nepa 7G)i> Tptdn/ roi'toi' koX i)u.i.TOviov in\ to o^v, outc avUrai tov ^oipiov toutov vXelov enl to /Sapu. Ov /otrje anaad. ye r} Ae'^i? tj Ka6' ei/ J^ fiopLOv \6yov TaTTOfj-evri, eVl Trjs avxTj? XeyeTat Tacreco?- aW yj ixev errt ri'js o^eia^, t] Se W errl tt^s /Sapeta?, tj 6e eir' 6.ix(f)0ii'. Tuji/ 5e afxtjiOTepa^ to.? raVei? ixovcritiu ai (leu Kara \ ftiav avWa^rtv avvs^6(ipii.ivQv exouo"*. t&» 6^e4To /3apu, as fit} n-eptajrw/aeVas ie.aXovfi.ev ^ / 6 ACCENTUAL PROSODY. But does it follow that our elocution is, therefore, monoto- nous 1 Far otherwise : it follows only that our accent carries with it no necessary change of note; and that the gradations of pitch which occur in speech are oratorical eflfects, not prosodial. As verse consists of words, and as words, again, consist of syllables, we have now to consider what are the laws which regulate the due sound of syllables in modern lan- guages ; in other words, what are the laws of accentual prosody. Of this prosody, which, first reduced to system by the Proven9al poets, has long prevailed in all modem languages, except the French,* the laws are fourfold : — - at Se iv ertpto re Koi erepw x<»pi$ e/coTepov €<^* eavrov rrfi' oiKeCav ^vAarrov ijyviTiv. Kai Tais fxev SKTvAAd/Soi? ovtey to Sia. fteVow xwptoj' /Sapvrijros re Kal o^vrrjTOS* Tats 6e 7roAu(TuAAa/3oi?, olat iror' av maiy, tj toi/ b$vi> tovov exovtra fiia iv ttoAAois /Sapetais eve{* "e-ited:! 1^ l*\ says Beauz^e, Tart de la prosodie existe par rapport h. notre langue, ,^i f ^ puisque nous en admirons les effets dans un nombre de grands ecrivains '^ 1| * iv dont la lecture nous fait toujours un nouveau plaisir ; mais les principes ^ \ ! NJ I n'en sont pas encore redigds en systfeme ; il n'y en a que quelques-uns ^pars ^"^ V ^k et Ih ; et c'est peut-etre une affaire de ge'nie de les mettre en corps.' — ,^ > \ * C'est en vain,' says another French writer, 'que quelques lexicographes ^ ^ ont voulu nous donner des regies certaines sur cette matifere : leurs efforts n'ont pas eu de succfes.' (* Dictionnaire raisonnee des diflScultes de la langue fran^aise.' Laveaux. Paris, 1818). The Abbe d'Olivet, indeed, tells us that French syllables are * long ' and ' short,' and classifies them accordingly ; but as he does not furnish a tangible definition of the alleged length and shortness, all he says on this part of his subject may be passed by as irrelevant. (See ' Prosodie fran^aise '). ■ '* * It is specially with reference to this law that the language of France ^^ ^ differs from that of every other European nation, as is pointed out (and I ^^ ^ believe, quite correctly), by Sismondi. What he says is in substance this : "N. ^ the Provengal prosody is adopted in all modern languages, except the , X-' French: all of them, with that exception, have in each word some one ^ ^ syllable on which the chief effort of pronunciation falls ; the syllable is thus ^i I i accented ; and it is the play of accentuation which gives harmony to verse : it >• the French, being strangers to this law, are disposed to deny its existence, ^»NI „ and in composing their own verse, are careful only to count syllables and ^1. X observe rhymes. (Lit. du Midi de I'Europe, vol. i. pp. 107, 110). ^i ^ Since writing as above, I have, however, observed that M. Brachet, in "i , his * Grammaire historique,' takes a different view ; asserting (pp. 72, 140) «k J \ that there does exist in French a fixed tonic accent, present always, he says, >^ ♦V in one of two places ; either on the final syllable of words having what is » /^^^ called a masculine termination, as in 'chanteur,' ' aimer,' 'finir,' or on the \^ penultimate of words having a feminine, that is, a mute termination, as m< ^^^. * sauvage,' ' verre,' 'porche ; ' and mute finafvowels being mere ocular effects, it comes practically to this, that the tonic accent, according to M. Brachet, ^ /falls on the last sounded syllable of ailj'rench words. 2y / Nothing caii b'e Simpler than sucE a law ; but there is this to be said .^ X against it, that no French prosodian anterior to M. Brachet appears to have ^*^ f been aware of its existence. , ^^<»*- The Abbd d'Olivet raises the direct question whether in French words "^ ^ any one syllable be more than any other distinguished necessarily by ^'H prosodial accent : and not onh' does he shrink from giving an affirmative > answer, but owns the subject of French accent to be a labyrinth, into which, X for fear of losing himself, he would rather not go too far. ^ > Are we, then, to infer that Academicians of the last century did not ^ ■* know how to pronounce French ? or has the pronunciation so much Si changed since their time ? V t ^^ monosyllables, however, sound equally forcible when taken by O THE TONIC ACCENT. 4. Accent belongs to syllables, not to theii* component parts, whether vowels or consonants. The first of these laws is not likely to be disputed ; but if any one be inclined to dispute the second, I would ask him to pronounce the following words, (or any others he may choose), and mark if, in so doing, he can give to any one of them a secondary accent without producing an effect which strikes him as peculiar : Pen'dant, depen'dant, indepen'dant, indepen'dantly ; Ar'biter, ar'bitrament, ar'bitrarily ; # / En'ergy, energe'tical, energe'tically ; "i^^^^^*^ Au'thor, autho'rity, autho'ritatively ; j^JvW. Pusilla'nimous, pusillani'mity. jr In every word, without exception, there will be found to be one syllable, and one only, on which the chief effort of pro- nunciation rests ; and the sound resulting from this effort is the Tonic accent.* themselves : thus, there is no difference between the noun substantive ' inn,' and the preposition ' in ; ' between t'ne article ' an,' and the name ' Anne ; ' between the preposition ' but ' and the verb ' butt ; ' between the particle ' to' and the numeral ' two ; ' but let any of them be sounded in conjunction with other words, and a marked difference is at once apparent. * At this point it may be well to take some account of the Latin acute accent. Cicero says, ' Ipsa natura, quasi modularetur horainum orationem, in omni verbo posuit acutam vocem, nee una plus nee a postrema syllaba citra tertiam.' (Orat. 18.) Quintilian's words are :' Apud nos brevissima ratio ; namquein omni voce acuta intra numerum trium syllabarum continetur, sive hse sunt in verbo solas, sive ultimae, et in his aut proxima extremae, aut ab ea tertia. . . . Est autem in omni voce utique acuta, sed nunquam plus una nee unquam tdtima, ideoque in dissyllabis prior. . . . Ea vero quae sunt syllabae unius erunt acuta aut flexa, ne sit aliqua vox sine -acuta.' (L. i. 5.) Here it is to be observed that Cicero uses the term ' vox,' (whfch half a dozen times in his context means 'voice'), to signify 'accent;' while Quintilian uses the terms ' verbum ' and ' vox ' as synonymous in the sense of * word : ' further, we may observe that Cicero treats the ' acuta vox ' as a ^omething which nature itself prescribes in speech ; hence, as natural laws are the same always, we seem entitled to infer that the thing he speaks of, (whatever it be), must exist in modern languages, as it did in the ancient. Xow, ancient authors nowhere define clearly how syllables were affected by acutation ; nevertheless, the passages above quoted establish beyond doubt the following collateral points : 1. Every Latin word has an acute accent : 2. No word has more than one acute accent : 3. The acute accent is always confined within the third syllable from the last : MONOSYLLABLES. 9 As regards monosyllables, tlie classification above given is wide enough to include all such as are of themselves abso- 4. It is never placed on a final syllable : 5. All words of one syllable are acuted. Thus we see that the acute and tonic accents, (whether essentially the same or not), have at least some properties in common : every word has one of each, and no word more than one : again, though Latin words, as we pronounce them, have no accent but the tonic, yet we always place it according to the rules laid down by Latin authors for placing the acute ; we never throw it farther back than the third syllable from the last, and we never accentuate a final syllable ; notwithstanding that the second of these rules is contrary to the usage of every modern language, and both of them to the usage of our own. As regards the rule that all monosyllables are acuted, we must under- stand Quint ilian either not to include particles under the term ' word,' or else to speak of them taken singly. There is good authority to show that by some ancient grammarians they were not deemed parts of speech, e. g. * Aristoteles duas partes esse dixit orationis, vocabula et verba.' (Varro, ♦ De lingua Latina.') ' Quibusdam philosophis placuit nomen et verbum solas esse partes orationis : csetera vero adminicula vel juncturas earum.' (Priscian, lib. xi. c. 11). Such is not, indeed, Priscian's own view ; but he holds, on the other hand, that though all prepositions and conjunctions, taken by themselves, are accented, yet that when used with other words, prepositively, they are not. ' Quo accentu pronunciatur " ab " (in " ab oris ") ? Per se acuto, in versu vero gravi, sicut et alias omnes prepositiones in suo loco positae, id est, prepositive,' (De xii. vera. ^n. c. 111^42.) Again, 'atque,' quem habet accentum ? Gravem in versibus, quomodo omnes prepositivas conjunctiones.' (Cap. v. 45). The 'Public School Latin Grammar' defines accent to be 'stress,' and svllables without accent to be ' barytone, having the grave, or we^sk tone, (^apus Toi/o?).' Thus the Latin acute accent is made identical with the modern tonic ; and, consequentially, the Greek, which all admit to be one in kind with the Latin, be the kind what it may. But the author does not tell us how to reconcile stress with short quantity ; and this we might expect him to do, unless he can show what means, other than stress, there is of marking long quantity. Nor ought it to pass unnoticed that here, by stroke of pen, he overrules the opinion of Dr. Foster, and many other scholars, who vehemently, and with much show of learning, if not of logic, maintain the 6^v»s and ^apus t6vo<;, (usually called acute and grave accent), to signify, not stress, or absence of it, but mere rise or fall in the musical scale. Lord Monboddo, indeed, affirms Dr. Foster to have made the matter !*o clear that nothing more need be said on it, since nothing but ignorance (his lordship thinks), or prejudice, could cause any one to be of a different opinion. (See ' Origin and Pro- gress of Language,' vol. ii. p. 252, and vol vi. 158). In the passage above-quoted from Dionysius, the words ' errl to 6^v ' and ' en-l TO /3apu ' do clearly, I admit, point to some change of note, upwards and downAvards ; but not clearly to such change by diatonic intervals, nor in necessary connexion with the acute accent: all that this passage can be said with certainty to do, is, to fix a limit, (namely, a fifth), beyond which the voice must not travel in speech. However, the controversy is not yet closed ; for Mr. Gladstone, I observe, defines the ancient accent to be musical pitch ; and protests against con- founding it with emphasis, or (as I should say) stress. (See report of speech before the Archaeological Society, 'Times,' June 9, 1877.) 10 MONOSTLLABLES. lutely entitled to accent ; but in every language there are some, and in ours a good many, monosyllables, which are accented, or not, according to the place they occupy in sen- tences, or to certain conditions which give them a gi-eater or less significancy. English monosyllables of this class are pronouns personal, possessive, and demonstrative, auxiliary verbs, signs of moods and tenses, and of comparison, and the negative particle * not.' The nominative of the personal pronoun is never ac- cented, when, without antithesis, it comes immediately be- fore or after, the verb : but always, and of necessity, there is accentuation, whenever antithesis has to be expressed through this, or any other, monosyllable not invariably accented. The following lines afford examples of the personal pro- noun accented, or not accented, according to these condi- tions : — How died he ? Death to life is crown or shame. All by him died, thou say'st. By whom died he ? * * Should the reader be inclined to consider this a case of emphasis, I must remind him that emphasis is an optional eflfect ; whereas what we here have is an effect by no means optional. There may be emphasis here, or there may not, but stress there must be, and necessary stress is accent. And this holds good as regards all cases where antithesis is marked by any monosyllable of the class above mentioned. In "the lines, — Pleas'd thou' shalt hear, and thou alone shalt hear — , Pleas'd thou shalt' hear, in spite of them shalt hear — , ('Art of Reading,' vol. ii, pp. 270, 271.) Sheridan gives examples of what he calls emphasis on words not usually forcible, namely, 'thou' and 'shalt;' but these words, though necessarily here accented, are not necessarily emphatic : the full meaning intended can be conveyed without either unusual enforcement of accent, or any change of note, though both may be used with good effect. In Harris's ' Hermes ' there is a passage aptly bearing on this point : *When we say," give me content," the " me " in this case is a perfect enclytic ; but when we say, "give me content, give him his thousands," the "me" and "him "are no enclytics ; but, as they stand in opposition, assume an accent of their own, and become true hfidorovovti.ivot., that is, rightly ac- cented.' Lindley Murray says that emphasis, in some cases, changes the seat of accent, and he gives, by way of proof, the following example : — ' He shall in'crease, and I shall de'crease.' But is there emphasis here ? does the change alleged take place ? As a test of the fact, and by way of counter-proof, I, too, propose an example, — ' While one increases, others are decreasing.' Now, a rule which is good for the future tense is good also for the pre- MONOSYLLABLES. 11 The nominative ' he ' of the first line, and ' thou ' of the second, are unaccented, while the nominative * he ' of the second line, being in antithesis, is accented. If other words intervene between the pronoun and the verb, the former will be accented, or not, according to the greater or less interval by which it is separated from the verb. ^N^ Again, oblique cases of the personal pronoun are not ^ accented when, without being followed by the relative, they ^ < come immediately after the verb, or after a verb, followed by a preposition (whether monosyllabic or not), as, for example, in * we heard him,' * they followed her,' ' he followed after me,' ' they spoke concerning her ; ' but if they go before the verb, separately, thus, * him we heard,' ' her he answered,* J they are ; and yet, not so, if preceding the verb, they are m themselves preceded by a preposition; as in the phrases * following, ' these words to him she said,' * after me they followed,' * concerning her they spoke.' Thus much with respect to the personal pronoun, when alternately requiring and rejecting accent. Possessive pronouns are not accented, save when standing alone : auxiliary verbs, and signs of tenses, are not accented, save when either standing alone, or separated by a con- siderable interval from the word they qualify. I i ■j 4 sent, and ' vice versa,' afl&rmatively and negatively : if, then, we see, at a ^ ' s.-^ glance, as we surely do, that the accent of ' increases ' and of 'decreasing' / ^^ cannot be displaced, the inference is unavoidable that there can be no such ^ X^ displacement in the examples given by Lindley Murray. <^ This writer has, no doubt, the authority of Ben Jonson, who says that when words like 'sociable,' 'insociable,' 'tolerable,' 'intolerable,' are used in opposition, then, but not otherwise, the seat of accent is on the syllable which points the difference. But let any one tr}' to pronounce these words on such a plan, that is, by throwing the accent on the first syllable of ' in- sociable,' ' intolerable,' as in ' pe'remptory,' ' pe'i'emptorily,' and reducing the remainder to an exact equality as regards absence of accent, and he will find, firstly, that the tongue iU responds to the attempt ; secondly, that the effect is far from pleasant. The right way of marking comparison between words of this class, is to pronounce the differential syllable of the second word in a slightly higher musical tone ; but if, by chance, we do allow ourselves at all to dwell on it, then we must remember that the stress thus used is an oratorical effect, to be permitted only so far as it is kept in strict subordination to the toni»» accent. 1 2 MONOSYLLABLES. Demonstrative pronouns are unaccented when imme- diately preceding a noun-substantive accented on its initial syllable ; as in the phrases ' that ancient building/ ' this object,' ' those persons ' ; but they are accented, or may be, whenever the noun is unaccented on its initial syllable, and the pronoun itself, preceded by an unaccented syllable, is followed, in the sentence, by the relative, expressed or under- stood ; as in Of that' forbidden tree, whose mortal taste — , My mansion is, where those' immortal shapes — . Again, the signs of comparison, ' more ' and ' most,' are not themselves accented, if their adjective have accent on its initial syllable ; but they seem to be so under other condi- tions.* Lastly, the negative particle ' not ' is never accented when it follows a verb, as in 'I heard not,' ' he discerns not,' ' they tarry not ' ; in nearly all other cases, it takes accent. English monosyllables, then, which never, under any cir- cumstances, take accent, are, — 1. The articles definite and indefinite ; 2. All particles ; f 3. All prepositions. J * See, however, as regards this point, p. 45. t ' W is not a particle when used to illustrate a hj^othetical proposi- tion, as in the lines following, — As if upon a well proportion'd dome — , As z/ Religion were intended — ; and in such cases it is accented. ' Now ' and ' then' are both clearly particles of connexion whenever they occur in the fourth sense ascribed to them respectively by Dr. Johnson. The essential dift'erence between the particle ' then ' and the adverb of time, is often, however, disregarded by poets. At times, it is not easy to decide what part of speech *but' is. This much, however, may safely, I think, be said : that whenever the word is used in the sixteenth sense pointed out by Dr. Johnson, namely, in the sense of ' without,' * had not this been,' as shown in the examples, — But for her native ornament of hair — , And but for mischief, you had died for spite — , To such mj' message is, and but for such — , it is adverbial, and takes accent accordingly. X It is necessary, however, at times to distinguish. ' Through ' and ' o'er ' (for ' over ') have a variable character : * through ' governs a case if ' QUANTITIVE ' AND ' ACCENTUAL.' CONFUSION. 1 3 That accent belongs to syllables, and not to letters, seems to me self-evident : still, it is a fact to be recognised that others think differently.* And here I will endeavour to point out wherein consists the difference between accentual syllables and quantitive, in verse : a subject which has caused much confusion among EngUsh writers ; some contending that our verse is formed on accent ; some that it is formed on quantity ; some that it is formed partly on accent, and partly on quantity ; some that it is partly, indeed, formed on quantity, yet that accent only need be attended to in the formation ; some that ac- cented and unaccented syllables correspond with long and short ones, without, however, being long and short cor- respondingly ; some that our syllables become long by posi- tion ; some that they are not affected by position ; some that they are always long if accented; some that they may be short, although accented; some that they may be long though not accented ; some that, failing accent, they are the idea in view be of mere passage between limits, or of means whereby a thing is done ; and * o'er,' if the idea be of mere position over an object, or of passage across an interval : but ' through ' is an adverb to express entire penetration, and *o'er ' is one to express entire covering of an object's sur- face, or, besides passage across an interval, arrival at the other sid . * It is held by Sheridan, and his followers, that in words like * all,' 'laid,' 'bide,' 'cube,' ' rood,' the seat of the accent is on the vowels ; Avhile in words like 'add,' 'led,' ' bid,' ' cub,' 'rod,' it is on the consonant ; and that according as it is on a vowel or a consonant, the syllables are ' long ' or * short ; ' length being marked by dwelling on the vowel, brevity by giving a smarter stroke to the consonant. But the distinctions here drawn are wholly referable to the differences of sound which exist in all vowels ; and mere vowel sound has no effect whatever on the accentuation, or quantity, of syllables. Whether we dwell on the vowel, or give a smarter stroke to the consonant, the effect is still syllabic sound intensified ; and any one in- tensified sound may be prolonged" with as much ease as any other. When Cicero tells us (Orat. xlviii. 159) that the first letter of 'in- clytus ' and of 'inhumanus ' is sounded 'brevis,' and the first of 'insignis,' ' infelix,' sounded ' producta,' he does not mean that the seat of quantity is on any of the letters named, for he knew the initial syllable of all four words to be long : manifestly, his meaning is that in the two first mentioned words the initial vowel has a curt sound, and in the two others, its primary sound. And here, be it observed, we find a plain contradiction of Sheridan 3 rash statement that length by position was always marked in Latin by dwelling on the vowels (Art of R., vol. ii. p. 9). As, then, in the ancient tongues, the seat of quantity is on the syllables, quite irrespective of vowel- sound, so on the syllables, and irrespective of vowel-sound, is the seat of accent in modern tongues. 14 CONFUSION, always short, and yet, by virtue of it, are shortened con- stantly.* Quantitive syllables are measurable by a fixed standard, and bear to each other fixed relative proportions : all long * Thus, according to Dr. Foster, a word like * distinguishing ' has its antepenultimate syllable long, and the rest short ; according to Mr. Mitford and Professor Newman, it has all the four long ; according to Sheridan, Walker, Dr. Nares, (' Elements of Orthoepy *) and Professor Blackie, it has all four short : Distinguishing, Distinguishing, Distinguishing. According to Dr. Foster, a word like ' testify ' has its antepenultimate long, its penultimate and final short ; according to Dr. Nares, Walker, and Professor Blackie, it has its antepenultimate and penultimate short, its final long ; according to Mr. Mitford and Professor Newman, it has its antepenul- timate and final long, and its penultimate short ; according to Sheridan it has all three short : Testify, Testify, Testify, TSstify. According to Dr. Foster, a word like ' semblance ' has its penultimate long and its final short ; and one like * content,' its penultimate short and its final long ; according to Mr. Mitford and Professor Newman, each has both long ; according to Sheridan, Walker, Dr. Nares, and Professor Blackie, each has both short : Semblance, content. Semblance, content, SSmblSnce, C(5nt6nt. According to Dr. Foster, words like 'banish,' ' clement,' are long on their penultimates and short on their finals ; according t© Mr. Mitford and Pro- fessor Newman, they are short on their penultimates and long on their finals ; according to Sheridan, Walker, Dr. Nares, and Professor Blackie, they are short on both : Banish, clement, BSnish, clement, BSnish, clgmgnt. According to Dr. Foster, words like * conclave,' ' umpire,' have one syl- lable long and one short ; according to Walker, Dr. Nares, and Professor Blackie, they have one short and one long ; according to Mr. Mitford and Professor Newman, they have both long ; according to Sheridan, they have both short : ConclSve, umpire. CQnclave, timpire. Conclave, umpire. CSnclSve, Umpire. According to Dr. Foster, Mr. Mitford, and Professor Newman, words like « fence,' ' bulge,' ' germ ' are long ; according to Sheridan, Walker, Dr. Nares, and Professor Blackie, they are short : Fence, bulge, germ, Fgnce, biilge, g6rm. So much do English authors differ among themselves when they apply to our language the terms of an inapplicable prosody. ANCIENT AND MODERN HEROIC. 15 are equally long, all short equally short : * one long is the equivalent of two short, two short of one long. Syllables distinguished by accent, or absence of it, ai© not measurable : there is no standard to which they can be referred ; and they bear no definite relation to each other. How much the two kinds differ in their effect on verse is shown by comparison of an ancient metre with a modern. The Latin hexameter has a variable number of syllables, invariably divided into six ' feet' ; each foot having not more than three or less than two syllables, the fifth having three, the sixth two. Thus the verse-syllables will vary from seventeen to thirteen, according as dactyls or spondees pre- vail in the four first feet ; and amid this variety of syllables, an exact gross equality of time is preserved between each foot and verse, f _ * Some writers among us, relying on certain well-known passages of Dionysius and Quintilian, contend that there are various degrees of length and shortness in sj'Uables ; and, on this principle, have drawn up tables, after the manner of Dionysius, showing the differences which exist between long as compared with long, and short as compared with short ; but they do not furnish, nor do the ancients, any standard whereby to fix the greater or less degree, nor point out what effect, on the structure of verse, these al- leged differences can have. The sole tangible standard of comparison be- tween quantitive syllables is proportion of two to one and one to two : (' Longam esse duorum temporum, brevem unius, etiam pueri sciunt.' Quintl. ix.) if degrees of quantity not referable to this standard exist, they cannot, at any rate, be brought to measurement ; and unless they can be shown to affect rhythm, we are entitled to disregard them from the rhythmical point of view. t The same exactitude is, however, not apparent in some metres : hence, if I were asked on what principle these are constructed, the answer would be that I do not clearly understand. Cicero, indeed, tells' us, with a directness not to be misunderstood, that some metres, and chiefly those of the best lyric poets, seemed little else than bald prose, unless they were chanted ; and that tragic lines had very much of the same appearance, unless the flute-player were by to set them off: as for the comic senarii, they were often, he says, so abject, that you could scarce detect in them either metre or rhythm. ' In versibus res est apertior : quamquam etiam a modis quibusdam cantu remoto soluta esse videatur oratio : maximeque id in optimo quoque eorum poetarum qui AupiKot a Graecis nominantur ; quos cum cantu spoliaveris nuda paene remanet oratio. Quorum similia sunt quaedam apud nostros, velut iste in Thyeste : Quemnam esse te dicam V Qui tarda in senectute — , et quse sequuntur ; quae nisi tibicen accessit, orationis sunt solutae similia : at comicorum senarii, propter similitudinem sermonis, sic ssepe sunt abjecti, ut nonnunquam vix in eis vel numerus aut versus intelligi potest.' (Orat. Iv. 183, 184. See also Quintl. Orat. 1, iv. 4, and 1, x. 29 ; and Dionysius, Sec. xxvi.). 16 ANCIENT AND MODEEN HEROIC. Onr heroic verse never varies from ten syllables (or, with a double ending, from eleven) ; it has no feet of definite dimensions ; and equality of time is not needed be- tween verse and verse. That it has not feet of definite dimensions is shown by this : feet are divisions of verse analogous to the notes in- cluded between beats of time in musical composition : for all rh}i;hm is divided by intervals which, equal or unequal, as regards number of notes or syllables, are quite equal as re- gai'ds dm-ation of time ; and the beats which mark the time of verse will always be found to depend on forcible syllables occurring in certain positions : now, some of our heroic verses have five, some four, some thi-ee, some only two forcible syllables ; whence it follows that as the proportion of beats to syllables, so will be the proportion of time allowed to the pronunciation of the syllables which have to be brought in between the beats;* so that, according to circumstances, two, three, four, or even five weak syllables may have to be uttered in exactly the same time as that which, in other cases, is allowed to but one. f * A theory has been put forward, with much confidence, by Mr. Steele, (in his ' Prosodia Rationalis '), and, following him, by the Rev. James Chapman, (in his ' Music of Language'), that the Latin heroic verse is not really hexameter, but octometer, and that English decasyllabic verse is at leasf hexameter, and often octometer. ' Whoever would pronounce,' they both say, * our heroic lines of ten syllables with propriety, must allow at least six cadences, by the assistance of proper pauses, to each line, and frequently eight.' Though it may be thought worth while to take notice of such opinions, few, perhaps, would think it worth while to discuss them. On the other hand, it seems to be assumed by the whole remaining body of English writers on versification, that our heroic verse consists of five feet necessarily. This again I hold to be a radical error, and the parent stock of many other errors. A foot, if it be anything, is a rhythmical phrase ; and there can be no rhythm where there is no forcible sound. ' Numerus in con- tinuationenullus est : distinctio, et sequalium ac saepe variorum intervallorum percussio, numerum conficit ; quem in cadentibus guttis, quod intervallis dis- tinguuntur, notare possumus, in amni prjetereunte non possumus.' (Cic. de Orat. 1. iii., cap. 48). Whoever adopts the theory of five feet for each verse will either be obliged systematically to attribute accent where none is due, or else to adopt some other expedient equally abnormal. In fact, the term ' foot,' being a quantitive terra, cannot suitably be ap- plied at all to accentual verse ; and it will not be so applied in this treatise. •j- James Harris remarks that music has five difierent lengths of notes. QUANTITY AND ACCENT. 17 From the same cause it follows also that this metre does not require, nay cannot have, equality of time between verse and verse : for if the beats vary from five to two, the time taken in reciting lines having less than five will be one, two, or three fifths, respectively, less than the time taken in reciting those which have that number. But whatever difierence there be between quantity and accent,* as they affect the structure of verse, there is found to and language only two of syllables : ( * Discourse on Music, Painting, and Poetry,' ) to me it seems evident that if the time given to the utterance of syllables, according to the ever-varying conditions which occur in measured /, speech, were reduced to musical notation, the differences between them would - - ^* appear to comprise all those existing between a minim and the smallest /^ ( imaginable sub-division of a quaver. * ' Though these two terms, quantity and accent, are in continual use, and in the mouths of all who treat of poetic numbers, I know of none to which more confused and erroneous ideas are attached ' (Thomas Sheridan, ' Art of Reading,' Vol. ii. p. 6, ed. 1776). No writer has done more than Sheridan himself to cause confusion of ideas touching these very terms. Our poetic feet, he says, are regulated by accent only,' and in composing English verse, a poet need not give the least attention to quantity ;2 notwithstanding that through neglect of it the melody of verse may be impaired : ^ in reciting verse, however, due attention to quantity is essential ;* but exactness in this respect is not essential,^ be- cause of the thing to be observed, which is constantly varying,^ there is no criterion : 7 accent is stress ; ^ and we have no means of marking long sylla- bles save by stress ; ^ yet accent itself may be a means of shortening syl- lables : ^'^ but though all accented syllables are not long, all unaccented ones are short;" except, indeed, when they become long by interjection of pauses : ^^ moreover, our verse, though formed wholly on accent, is formed on quantity likewise ;'3 for we have duplicates of each foot,'^ some accentual, some quantitive, available, at pleasure, for different purposes -^^ and accord- ing as the accent is seated on a long or a short syllable, these feet produce different effects -^^ and the way of reducing these duplicates of feet to equal- ity of time is by rests, as in music ; the larger^^ proportion of pauses, compen- sating for the' smaller proportion of sounds ;i8 so that we find united, in English verse, all the powers of wind and stringed instruments : accented syllables are necessarily long, or necessarily short, according to the quality of the vowel sound '^'■^ long, for instance, if the sound be close, as in ' holy ' and ' eastern ; ' short, if it be open, as in ' holly ' and * western ; ' and length, in the one case, is marked by dwelling on the vowel, brevity, in the other, by giving a smarter stroke to the consonant i^o not but what we may dwell on coiiSonants,2i making long thereby sounds necessarily short j at least, we may » Vol. ii. pp. 5, 29, 32, 34, 77, 87, 162. " Ibid. pp. 29, 77, 166. ' Ibid. pp. 91, 95, 308. * Ibid. pp. 29, 291. = Ibid. p. 77. « Ibid. pp. 26, 27, 90, 162, 277. ' Ibid. p. 78. « Vol. i. pp. 104. 108, 110, 121 ; vol. ii. p. 32. » Vol. ii. p. 15. "• Ibid. pp. 8, 9, 10, 23, 24, 26, 162, 209. " Ibid. pp. 26, 162. " Ibid. 212. " Ibid. pp. 64, 65, 66, 223, 233. " Ibid. pp. 62, 63, 64, 65, 66, »8, 89, 166, 229. " Ibid. p. 62. '* Ibid. pp. 63, 164. " Ibid. pp. 166, 223. " Ibid. p. 166. " Ibid. pp. 8, 9, 10. ■"> Vol. i. pp. 121, 122 ; vol. u. pp. 9, 10, 23, 24. " Vol. i. pp. 97, 98 ; vol. ii. p. 212. G 18 COMPARISON. be none between them as they aiFect the sound of separate syllables or words, whenever the comparison be made under well defined equal conditions : * long syllables claim neces- sarily to be enforced ; short, to be passed over without en- forcement ; f thus we mark syllabic proportion by doing, or not doing, the very thing in which accent, or absence of it, consists. in monosyllables, and final syllables of other worris, but nowhere else : thus, we may say ' swell the proud note,' ' fulfil your purpose ; ' but must not say the ' swell-ing note,' ' fulfil-ing all ; ' because this would be to transgress a fundamental law, by severing syllables from words ; ^ (for though to dwell on a long accented penultimate, as in ' ho-l}',' does not cause severance, yet to dwell on a short one, as in * holl-y,' does) ; and yet, in cases of emotion, the law, no doubt, may properly be transgressed : 2 but after all, there is nothing in the nature of any sounds to render them necessarily either long or short, as their quantity depends on our dwelling on them or not ; ^ and although the nature of some sounds does, indeed, as was admitted, require that we should dwell on them,* yet we are not bound to do so unless we like ; ^ the fact being, that when words are arranged in sentences, the quantity of their syllables depends on the relative importance of their sense ;^ in short, it is emphasis which regulates quantity ;7 that is, the ac- cented syllable of each emphatic word acts as a standard of measurement, and, like a key-note, gives the tone and proportion to the syllables of the remaining words.** Such is the substance of his utterances on points which others, he thinks, have not clearly enough explained. If Mr. Mitford's utterances on quantity and accent were handled in the same way, they would scarce present a better appearance. * For instance, a word like ' contendunt,' consisting of three long syl- lables, cannot be compared with one like ' conte'nding ; ' because here two long syllables would be compared with two unaccented ; nor can a word like * SnimS,' consisting of three short syllables, be compared with one like * a'nimal ; ' because here a short syllable would be compared with an accented one ; but between words like ' carminS, ' and ' ha'rmony,' ' sSreuS, ' and * sere'nely,' 'inimicS,' and * inimi'cal,' that is, between words in which long syllables correspond with accented, and short with unaccented, throughout, tiie conditions are equal, and the effect found to be identical. f It may be gathered, from these remarks, that I impugn our customary manner of pronouncing Latin hexameter verse. In doing so, however, and throughout my remarks, I assume two things : firstly, that such verse is composed of syllables which, bearing fixed proportions relatively to each other, are arranged in feet neither more nor less than six ; secondly, that it oixght to be read in such a way as to render perceptible the prosodial quality of each syllable, and the regular ictus of each foot. And if the first of these ])ostulates be granted, but not the second, then I should ask, for which is verse meant ? for men's eyes or for their ears ? and if for their ears, then to what purpose is it built on a particular system, unless we have audibly presented to us the effects with which by that system it is accredited ? Now, the usage of modern Europe is, we know, to pronounce Latin verse ^ Vol. 1. pp. 98, 99. =» Ibid. p. 99. ' Vol. ii. pp. 90, 91. * Ibid. pp. 8, 9, 10. * Ibid. p. 91. « Ibid. pp. 27, 163, 270. ' Ibid. 27, 163, 270, 286. * Ibid. pp. 163, 291. A QUESTIONABLE ASSERTION. 19 Dr. Johnson has twice said * that one language cannot communicate its rules to another : on the contrary, I hold according to accent, as defined in this treatise ; a usage which entirely ignores the laws of quantity as we have been taught to understand them. If this be right, then either the terms ' long ' and ' short ' must be con- fessed not to support the meaning theoretically ascribed to them by ancient authors, or else Roman poets composed verse by one system of prosody and pronounced it by another. And this latter supposition seems adopted by Lord Karnes, who, in his ' Elements of Ci'iticism ' (pp. 94, 95) boldly asserts dactyls and spondees to be artificial measures, invented but to try the accuracy of composition, and so little serving any other purpose, that were we to pronounce according to these feet, the melody of the metre would be destroyed. But it is hard to understand how accuracy of composition could be attained by mixing ingredients which have no effect on the compound. To me it seems that Latin hexameter verse, read according to accent, has a sound which, if rhythmical, is hardly metre ; or, if metre, then of a kind as different from the kind due as any one thing can well be to another. And yet all admit that in this same verse, read according to quantity, there is a something which strikes us as artificial. Are we sure, then, be- yond all doubt, that the ancients did so read it 'i being sure, why do we not follow the known right course ? But if, in repeating such passages as Quae circum litora, circum Piscosos scopulos, or Trojae qui primus ab oris, Quintilian pronounced (as he tells us) ' circum litora,' and ' qui primus' (and for consistency's sake he should have added ' circum piscosos '), each as one word, without any pause (' tanquam unum enuntio, dissimulata distinc- tione,' 1. i. 6, 7), would not the method which these examples indicate render impossible that correspondence between metrical feet and beats of time in which reading by quantity consists ? Again, when the same author, speaking of the poetic feet which occur in oratory, warns his readers not to be too careful in measuring feet and syl- laoles, and points to versifiers, who regard (he says) the general flow of their periods, not the five or six parts of which verse may consist (' neque tam sunt intuendi pedes, quam universa comprehensio, ut versum facientes totum ilium decursum, non sex vel quinque partes, ex quibus versus constat, aspi- ciunt,' Orat, ix. 4. 116), would it not from these words seem as though verse, read naturally, were apt to outgo its bounds, and confound the distinctions between feet ? Nor is the naturalness of poetic feet proved by showing that they exist in oratory : they do, no doubt, on the prosodial data ; but on these it is that the question turns : we cannot speak otherwise, Quintilian says, than in long and short syllables of which feet are made ; (ix. 4. 61,) and there it no foot, says Dionysius, which, found in verse, is not also to be found in prose: (sec. xvii.,) but unless the true customary utterance of syllables in prose were such as to render their length or shortness perceptible in the sense ascribed, then the versification which assumes them to be long or short in that sense must be held to rest on an artificial basis. How, then, shall we decide ? how adjust the strange discrepancies between practice a^id theory which so much perplex us in this matter ? I do not, for my own part, pretend to say : all I do is to state a dilemma, and ask what escape there is out of it. * ' Kambler,' No. 88, and ' Life of Milton.' c 2 20 COMMON PROSODY, LAWS IN COMMON. that rules of verse are necessarily common between languages which have a common prosody : for verse being the musical element of poetry, affecting the ear by a certain modulated arrangement of words, it seems to me that the laws which rule the sound of syllables in any two languages being the same, a similar arrangement of equivalent syllables in each must needs produce a similar effect : * more or less pleasing the effect will be according to the better or worse quality of the sounds ; f but as in music a melody remains the same melody on whatever instrument it be played, or to whatever words it be set, so verse may please more in one language than in another, and yet be in both essentially the same verse. If this be so, it remains but to fix some standard of excellence in verse, and then to investigate the laws on which it is founded. * ' It is impossible that the same measures, composed of the same times, should have a good effect on the ear in one language, and a bad effect in another.' (Goldsmith's Essays, No. 18.) Disagreeing with everything else in Goldsmith's Essay on Versification, I have, however, quoted from it this passage, as the principle here laid down is the same as that laid down in the text above, though the application is different ; Goldsmith's object being, to show that our language is not repug- nant to the Greek and Latin measures : but in arguing thus, he assumes the prosodial conditions to be the same on both sides ; whereas they are quite different, and so the argument fails. Whatever Goldsmith might think or say, he took care not to give us any English verses of his own, founded on the ancient metres. It is worthy here to be noted that Ben Jonson ( ' English Grammar ' ) and Dryden ( ' Discourse on Epic Poetry,' and ' Preface to Alboin and Albinus ' ) both expressed an intention to publish a treatise, proving the suitableness to our language of a quantitive prosody, like the Greek and Latin, and that both died, full of years, withovit carrying out the intention. The inference I draw is, that they found the task, on trial, less easy than they had imagined. f Whatever advantage some modern languages may have over ours, we must remember that as all (save the French) are subject to the same system of prosody, all syllables of any one are, ' caiteris paribus,' equivalent, for rhyth- mical purposes, to all syllables of any other. Now, equality of conditions here depends solely on the presence or absence of accent : hence, the final syllables of words like * subjects,' * pretexts,' encumbered though they be with three consonants, are equivalent, for the said purposes, to the corresponding syllables of Italian words like ' canto ' and ' bella ; ' and are equally with them, therefore, eatitled to be used in any position where unaccented syllables are required. If effects, comparatively inferior, result, these are due to that inherent meanness of musical quality which pervades the whole language, and which should be taken account of, at starting, by every one who sets himself to compose, or read, English verse. A bag-pipe is a poor instrument in comparison with the finest organ ; but if a man choose to play on, or listen to, a bag-pipe, he must be content to take such sounds as it can give him. A STANDARD TAKEN. 21 Now, in this matter, the human ear is beyond doubt the supreme judge.* If, then, in any language the melody of verse be such as to leave the most sensitive ear nothing further to desire, the verse of such language may safely be held to furnish the standard we are seeking ; nor will there be any difficulty in defining its laws, since the laws of such verse will ever be found uniform and consistent : indeed, were they not so, the ear's judgment might be questioned : for if the true test of good verse be that it satisfies a perfect ear, the true test of a perfect ear is that its judgments, if analysed, shall be found based on definite laws consistently applied, f Tried by this test, the vei'se of Italy seems perfect in structure : it may be taken then as our standard : but equally entitled to be so taken is any verse which with equal con- sistency observes the same laws. J * ' Musicians have taken pains to discover the principles on which con- cords and discords are to be arranged so as to produce the best effects ; and have brought the whole art of harmony within the compass of a certain number of rules, some of which are more, and some less, indispensable : these admit not of demonstrative proof ; though some of them may be inferred by natural deduction, from the very nature of sound : yet the supreme judge of their propriety is the human ear. They are, however, founded on observa- tion so accurate and just, that no artist ever thought of calling them in question.' (Beattie, 'Treatise on Poetry and Music.') f ' In versibus modum notat ars, sed aures ipsae, tacito cum sensu, sine arte definiunt.' (Cic, Orat. ix. 203.) It is quite possible for verse to be faultless, even though its author be un- able to define the qualities which make it so : but we need not, in such cases, suppose authors instinctively to have hit upon the system of versifica- tion they practise. In the country near Rome and Florence, it was common formerly to find young persons of either sex, who, not knowing how to read, ignorant of all naetrical laws, and guided, as we are told, by ear alone, could pour forth harmonious verses on any subject that might be proposed to them: (Metastasio : 'Note alia Poetica di Orazio,' v. 361), but these young peasants had been used from childhood to recite, and hear recited, the per- fectly correct verse of others ; and their ear had become perfected in consequence. X In Professor Craik's * English of Shakespear,' I find the passage follow- ing : ' The mechanism of verse is a thing altogether distinct from the music of verse : the one is a matter of rule, the other of taste and feeling. No rules can be given for the production of music, or the musical, any more than for the production of poetry or the poetical. The law of the mechanical con- struction of verse is common to verse of every degree of musical quality : to the roughest and harshest (provided it be verse at all), as well as to the smoothest and sweetest. Music is not an absolute necessity of verse ; there are cases in which it is not even an excellent or desirable ingredient. Verse is sometimes the more effective for being unmusical. The mechanical 22 LAWS OF ACCENTUAL YERSE. These laws are of two kinds : those which apply to all verse alike ; and those which apply to particular metres. Of the first kind are the following : — 1. The rhythm of verse depends on the occurrence of accent at certain determinate intervals : * 2. The accent must be none other than the strict prosodial one of each word : f law or form is indispensable. It is that which constitutes verse. It may be regarded as the substance, musical character, as the accident or oi'nament.* (P. 30.) That the mechanism of verse is distinct from its music may safely be granted, for the one is cause, the other effect : safely also may we grant that no rules can be given for the production of music : all I say is, that sounds, to be musical, must needs obey some rules which are to be given. But music, says Professor Craik, is not a necessity of some verse : wherein, then, I ask, does such verse differ from prose ? In the mechanical law, or form, we are told, and these are indispensable : but if these do not tend, more or less, to melody, to what do they tend ? and what use are they ? and how can useless things be indispensable ? * ' Numerosum est in omnibus sonis atque vocibus, quod habet quasdam impressiones, et quod metiri possumus intervallis aequalibus.' (Cicero, ' De Orat.' xliii. 83.) ' Distinctio, et sequalium et saepe variorum intei^vallorum percussio, numerum conficit.' (Id. 186.) f Mr. Mitford contends (p. 63) that many English words have more than one ' distinguishing ' accent : one, he says, is always predominant, and pro- perly called, by way of eminence, ' the accent ; ' but trisyllables may have two ' distinguishing ' accents, and longer words two, or more, according to their length. This theory cannot be applied without vitiating the natural rhythm of language. If the alleged accents exist they will be effective ; and the test of their effectiveness is whether or not they can be marked by beats of time. Now, an insufferable artificialness of utterance results from letting a beat fall on any syllable save that which carries the tonic accent. Professor Craik says that ' any syllnhie whatever,^ occurring in a place where accent is due, ' may be deemed accented for purposes of verse ; ' and verses formed on this principle he does not admit to be licentious, even while admitting them to conflict with ' the normal, natural, and customary rhythm of language.' (See ' English of Shakespear,' from p. 35 to p. 38.) But if such verse be not licentious, it were hard to say what licence is. The same author writes : ' In every language, the principle of the law of verse undoubtedly lies deep in the structure of the language. In all modem languages, at least, it is dependent on the system of accentuation established in the language, and would probably be found to be modified, in each case, according to the peculiarities of the accentual system ' (p. 30). If the law of verse, established in modern languages, be modified in any one of them on account of peculiarities alleged to exist in that one, those who modify the law, or sanction its modification, on such a plea, are bound to point out what are the peculiarities they mean. But with what relevancy do we talk about laws of verse being dependent on an established accentua- tion, if every syllable in the language may be made to carry accent, for verse-purposes, at pleasure ? HEROIC TERSE OF MODERN TONGUES. 23 3. No syllable not sounded in declaimed prose should be sounded in verse : * t /^ 4. A concourse of final and initial vowels tfCgh^ not occur ^A^*'^ without elision. | ^^ _A Elision is a process by which two vowels, meeting in separate words, are made to count but as one syllable. These four rules will be found on examination to form the basis of all vei*se in the Romance tongues, except the French ; but, proposing to treat fully of them farther on, I will now take in hand the metre known as the Heroic verse of modern languages. In its primary form, this verse consists of ten metrical syllables, each even syllable being metrically accented ; \ but the form varies constantly ; and there is no syllable which always must be so accented, except the tenth, and none which never can so be, except the ninth. By metrical syllables, I mean those only which count in verse ; and by metrical accents, those only on which beats of time depend. In Italian, and other Romance languages, the verse takes, with rare exceptions, an unaccented eleventh syllable after , / the accented tenth : § in English, this occurs but casually; ,. ? /« for the seat of accent in our words is far more favourable to ^ • single endings than to double. ^* 'i^ * Only so far as declamation makes a difference, should the sound of words in verse differ from the sound of the same words in common speech. The difference consists in a somewhat higher pitch of voice, a much slower time, a more strongly marked accent, and a more distinct utterance of unac- cented parts. * Plus la prononciatioa est lente,' says the Abbe d'Olivet, 'plus la prosodie devient sensible: on lit plus lentement qu'on ne parle; ainsi la prosodie doit etre plus marquee dans la lecture, et bien plus encore au barreau, dans la chaire, sur le theatre.' (' Prosodie Fran9aise.') Now recited verse is unquestionably declamation ; and the fact is never to be lost sight of. t Such is the general rule as regards final and initial vowels : but reasons will be given hereafter showing why exceptions may be made in certain cases. I The metre, then, may be termed ' iambic,' because primarily it consists of accentual phrases, each corresponding in sound to a quantitive iambus. § Verses even of twelve syllables are admissible, provided there be no accent after the tenth syllable ; if the twelith be accented, the verse becomes an Alexandrine. 24 PRIMARY FORM, AND VARIATIONS. Our heroic verse, then, may have metrical accent in a great variety of positions ; namely, — 1. On the 2nd, 4th, 6th, 8th, 10th syllables 2. „ 1st, 4th, 6th, 8th, 10th ; 3. „ 1st, 4th, 6th, 7th, 10th ; 4. „ 2nd, 4th, 5th, 8th, 10th ; 5. „ 2nd, 4th, 6th, 7th, 10th ; 6. „ 2nd, 3rd, 6th, 8th, 10th ; 7. „ 2nd, 3rd, 6th, 7th, 10th ; 8. „ 1st, 4th, 6th, 10th ; 9. „ 1st, 4th, 8th, 10th ; 10. „ 1st, 6th, 8th, 10th ; 11. „ 1st, 4th, 7th, 10th ; 12. „ 2nd, 4th, 6th, 10th ; 13. „ 2nd, 4th, 7th, 10th ; 14. „ 2nd, 4th, 8th, 10th ; 15. „ 2nd, 6th, 7th, 10th ; 16. „ 2nd, 6th, 8th, 10th ; 17. „ 2nd, 3rd, 6th, 10th ; 18. „ 3rd, 6th, 7th, 10th ; 19. „ 3rd, 6th, 8th, 10th ; 20. „ 4th, 5th, 8th, 10th ; 21. „ 4th, 6th, 7th, 10th ; 22. „ 4th, 6th, 8th, 10th ; 23. „ 1st, 4th, 10th ; 24. „ 1st, 6th, 10th ; 25. „ 2nd, 4th, 10th ; 2Q. „ 2nd, 6th, 10th ; 27. „ 2nd, 8th, 10th ; 28. „ 3rd, 6th, 10th ; 29. „ 4th, 6th, 10th ; 30. „ 4th, 7th, 10th ; 31. „ 4th, 8th, 10th ; 32. „ 6th, 7th, 10th ; 33. „ 6th, 8th, 10th; 34. „ 4th, 10th ; 35. „ 6th, 10th.* * I have not included, above, those forms in which the accentual move- ment is from the first syllable to the third, or from the fifth to the seventh, because in such cases the accent on the first and fifth syllable does not cause an additional beat of time. Nevertheless, it is to be remembered that accent in any position gives variety to verse, even when it has no effect on the time. VAEIATIONS. 25 There are, then, thirty-five forms possible of this metre, as regards metrical accent ; of these, eleven (viz., the 1st, 2nd, 8th, 9th, 12th, 14th, 16th, 22nd, 26th, 29th, and 31st) have all one property in common, which is, that no uneven syllable, save the first, is in any of them metrically accented ; but as they are all, more or less, in common use among our poets of past and present time, I do not deem it necessary to say anything further about them. The characteristic of the 3rd, 4th, 5th, 6th, 7th, 15th, 17th, 20th, and 21st forms is that in all of them either the third, fifth, or seventh syllable receives metrical accent concurrently with the second, fourth, or sixth, as is seen in these lines of Milton, — 2nd 3rd Inclines, here to continue, and build up here — , 4th 6th Amaz'd I stood, harass 'd with grief and fear — , 6th 7th Ease to the body some, none to the mind. Of such arrangements I have only to say that they seem to me quite unobjectionable; and that though pronounced by some, and especially by Dr. Johnson, to be vicious and of defective harmony, they occur, over and over again, in every page of the best Italian poets. * In the 18th, 19th, and 28th forms, none of the first six syllables, except the third and sixth, are metrically accented ; and in the 11th, 13th, and 30th, none of the last sixth, ex- cept the seventh and tenth. So very common, in Italian verse, i§ uneven accentuation of the first six syllables, that we need seek no other reason to account for Milton's frequent use of it ; "f but English poets. * Alone among our poets, Milton seems to have studied versification in the best school ; and varieties of measure, derived by him from that source, give a distinctive charm to his numbers. Still, he cannot be taken as a safe guide in verse : for though he copies from a perfect model, yet he copies capriciously, and without heeding the causes of its perfection : hence, while Romance poets will scarce have a faulty line throughout a whole poem, he, tried by their standard, has several in each page. f It proves, however, a sad stumbling-block to Milton's commentators. Thus, with reference to the line And Tiresias and Phineas, prophets old. 26 VARIATIONS. in using this form, would do well to remember three things : firstly, that the movement, being hurried, suggests the choice Dr. Newton remarks, * Dr. Bentley is totally for rejecting this verse, and ob- jects to the bad accent on "Tiresias ; " but, as Dr. Pierce observes, the accent may be mended by supposing the interlined copy intended this order of the words, And Phineas, and Tiresias, prophets old.' Again, with reference to the line Universal reproach, far worse to bear, he writes, * Here are two trochees, and not an iambus till the third foot ; and 80 likewise in v. 876, Through the infinite host, etc. — This measure is not common ; but, as Mr. Jortin observes, Milton often in- serts harsh lines when he could easil}' have altered them, judging, I suppose, that they have the same effect which discords have in music' Jortin remarks on the verse But to vanquish by wisdom hellish wiles — , * Milton lays the accent on the last syllable of " vanquish," as, elsewhere, in " triumph ; " and in many cases he' imitates the Greek and Latin prosody, and makes a vowel long before two consonants.' Sheridan, in his ' Art of Reading,' says ' it is an inviolable law that two homogeneous feet (dactyls and trochees) should not be found together in a verse.' Quoting then eight lines from Milton, where this happens, he pro- nounces them to be ' false metre ; ' adding, ' I have been surprised, in read- ing Milton, who was so perfect a master of numbers, to find so many lines that have not the least air of verse, and which would not have slipped from the pen of our middling poetasters.' .... 'In most of these you find there is not the least pretension to verse ; . . . . but there are other passages, in which he industriously started aside from all rules of metre, for the sake of a more vigorous expression ; as, for instance, in the following description of of sin opening the gates of hell to Satan — , Then in the key-hole turns The' intricate wards ; and ev'ry bolt and bar Of massy iron and solid rock with ease Unfastens. On a sudden open fl}', With impetuous recoil and jarring sound. The' infernal doors : — these lines are certainly,' he continues, ' exceedingly expressive, but can- not be called verse ; nor do I think so great a latitude is allowable.' Sir Egerton Brydges, in his ' Remarks on Milton's Versification ' (vol. v. ed. 1836), proposes to 'scan,' as I am about to mark them, the two first words of the line Universal reproach, far worse to bear — . Judging from this, and other indications, it must have been a pleasant thing to hear Sir Egerton Brj-dges recite Milton. Mr. Mitford, however, gives a qualified assent to such lines : * Perhaps, also, he says, ' the transfer of the accent in the second foot may be more freely permitted to dramatists than in epic verse. Milton evidently had some partiality for an extension of this license ; in imitation of the Italians, with whom this aberration of the accent in the first and second foot of the same verse is more common than in either of them alone : insomuch that Tasso, who is reckoned among the more scrupulously harmonious poets, has begun his most admired poem (the " Jerusalem Dehvered ") with such a verse. Canto I'arme pietose, e 11 Capitano.' VARIATIONS. 27 of smoothly moving words ; * secondly, that long words are here preferable to short ones ; thirdly, that a succession of six monosyllables is apt to have a very poor effect, especially when the first of the six is unaccented ; and this arises from the exceeding meanness of our monosyllabic particles, of which, in the case supposed, four would have to be crowded into one half line. Uneven accentuation of the last six syllables, though far less common than the foregoing, we find used, at times, with good effect ; as, for instance, by Dante, 7th E come ^ quel che con lena affannata — , 7th Riprese via per la piaggia deserta — ; and by Petrarch, 7th Se per salir aU'eterno soggiomo — , 7th Plena di rose e di dolci parole — ; and by Tasso, 7th Ch'ogni suo pregio 5 non fatto, ma nato — ; f He then quotes verses from Milton, and among them these two, Irrecov'rably dark, total eclipse — , Irresistible Samson, whom unarm'd — ; adding, that to * do them justice, the first syllable of each word must be made long by the distinct pronunciation of the double consonant.' But here I quite differ with him : I deny that these syllables have any title to be made ' long,' or, as I should say, ' accented,' or that, metrically, they have any need so to be : because (as I have said before) when the third syllable is metrically accented (the movement being from the first to the third, and from the third to the sixth), accent on the first syllable does not give an additional beat to the verse. * That is, words not over-encumbered, in their unaccented parts, with a concourse of distinctly sounding consonants. t ' Dante was not averse to such verses ; Petrarch seems to avoid them carefully ; Ariosto used them sparingly ; Bemi, in " Orlando Inamorato," has a great many ; but Tasso has not a single line.' (Baretti's ' Grammar of the Italian Tongue.') Experience has taught me that no assertion, made by the author of * La Frusta Letteraria,' is to be taken on trust. It is not true that Petrarch seems carefully to avoid this form ; for it occurs in his poems quite often enough to show that he had no dislike to it ; nor, again, is it true that Tasso has not a single line of this kind, as is shown by the one quoted above ; and others, doubtless, may be found. As regards Dante, instead of saying that he is not averse to such lines, it would be far nearer the truth to say that he shows a strongly marked partiality for them. Elsewhere, in the same work, this author speaks in disparagement of Tasso's versification, making against it specific allegations which have not a particle of truth in them. 28 VARIATIONS, and by Alfieri, 7th Prima tremenda giustizia di sangue — , 7th Che tu non pianto, ma sangue nemico — . Of the next coming examples, which are from Milton, the utmost, perhaps, to be said is that they neither please nor displease : — 7th Light from above, from the fountain of light — , 7th Things not reveal'd, which the invisible king — , 7th The Pontic king, and in triumph hath rode — , 7th Hereafter join'd, in their populous tribes — , 7th And dust shalt eat, all the days of thy life — , 7th And made them bow to the gods of their wives — , 7th Before thy fellows ambitious to rise. And here let it be understood that while holding un- evenly accented verses to be worthy, the one form more, the other less, of adoption in our language, I do not, however, recommend the indiscriminate use of them : I would not, for instance, venture to disturb, by their presence, that well- balanced stateliness of movement which seems to be the marked characteristic of our heroic couplet, and the alter- nately rhyming stanza of Gray's Elegy. The 10th and 24th forms appear in verses like Ha'rmony to beho'ld in we'dded pai'r — , Sca'ndalous or forbi'dden in our la'w — , Fe'cemi la divi'na potesta'de : and the chief feature to be noticed in them is the absence of accent between the first syllable and the siKth. Of the 23rd, 25th, 27th, 32nd, 33rd, 34th, and 35th forms the peculiarity is that all have, in some position or other, a sequence of five weak syllables (that is, accent is suspended either till the sixth syllable, or from the second till the eighth, or from the fourth till the tenth) ; while the VARIATIONS. 29 34th and 35th forms present the further peculiarity of having but two accents. The eflect of accent thus variously suspended is shown in the lines following ; 33. Irrevocabilme'nte il mi'o desti'no — , (Alf. Timol.) 09 /Infaticabilme'nte a'gile e pre'ste — , (Tass. G. L. 1. 9.) ' I Velocissimame'nte egli si spi'nge — , (Id. 7. 38.) oj rO To'sco che, per la citt^ del fuo'co — , (D. Inf.) * I Be pe'nitent and for thy fault contri'te — , (M. P. L.) 23. The'n with the multitude of my redee'm'd — , (Id.) 25. Se pri'a io ra'tto infaticabilme'nte — , (Alf. Br. Sec.) 35. Che dolcissimame'nte si diffo'nde — .* (Tass. G. L. 18. 18.) Thus much as regards metrical accent. f But accents, not metrical, are constantly admitted in verse ; nor is there * Though not able to furnish an example of the 34th form, I do not doubt that examples are to be found ; for it follows obviously from other forms ; especially from the 27th and 2ord. If we can pass from the second syllable to the eighth without intermedi- ate accent, we can pass without it from the fourth to the tenth ; and there is not even a ' prima facie ' difficulty in reaching the fourth syllable without previous accent. The verse, above qiioted, as an example of the 23rd form, is accented on the first, fourth, and tenth syllables ; but if we were to cast out the accented adverb 'then,' and use, instead of it, the unaccented con- junction ' and,' we should thus obtain a verse of the 34th form. f Some writers, as 1 have shown above (pp. 21, 22), say plainly that any syllable whatever, falling in a position where accent is due, may be deemed accented for purposes of verse ; others, less plain-spoken, act on the same principle ; for no one who adopts the theory of five feet for each verse can avoid using fictitious accentuation in every verse which has less than five tonic accents metrically arranged. Thus, after telling us, quite rightly, that all modern languages, except the French, are accented on one syllable only of each woi-d,and that it is the play of accentuation which gives harmony to verse, Sismondi proceeds to quote, and by way of illustration, to * scan,' in five feet, verses selected from Provenyal, Italian, Spanish, Portuguese, and English poems ; and lo ! straightway we find him contradicting in practice the very things he had before said ; accenting insignificant monosyllables, disaccenting significant ones, and attributing, at pleasure, double accent to words of three syllables and upwards. (' Lit. du Midi de I'Europe,' vol. iii., from p. 114 to p. 119.) Those who care to observe further how the theory works are referred to Lord Monboddo's ' Origin and Progress of Language ' (vol. ii. part ii. p. 326) ; to Sheridan's ' Art of Reading ; ' to Mitford's ' Harmony in Language,' from p. 89 to p. 99 ; to Sir Egerton Brydges's ' Remarks on Milton's Versifi- cation ; ' to Professor Craik's * Language of Shakespear.' One quotation will suffice : but here I must repeat that * a foot,' if it means anything, must have a forcible sound ; that the forcibleness of syl- labic sound always coincides with beats of time ; and that a marked arti- ficialness of utterance results from letting a beat fall on any syllable which does not carrv the tonic accent. 30 NON-METRICAL ACCENT. any limit (save that of the verse itself), to the number ad- missible. But if verse consists in an alternation, more or less varied, of weak and forcible sounds, what accordance, it may- be asked, with such a principle, can there be in an unbroken succession of sounds all forcible ? To this enquiry no satisfactory answer can be given save by reference to a law which has here to be stated : that where accented syllables, not separated by a pause, come to- gether, the second of any two always dominates over the first, and the third of any three over the second ; by which I mean, not that the fii-st or second syllable ceases (for it can- Now, in page 37 of his above-mentioned work, Mr. Mitford quotes from Shakespear five lines, three of which, like The crow doth sing as sweetly as the lark, have but four tonic accents, and, therefore, but four beats ; and the rest, like No better a musician than the wren — , but three. Nevertheless, they are all divided into five * feet.' By way of proving that Italian verse permits double accentuation in tri- syllabic and other words, the same author quotes Baretti (' a man of little talent, and little learning '), who, in his Grammar, seems to say that each even syllable in the line Che ritrovarsi in servitii d'amore, is accented. But if Baretti meant this, it only proves that he had learnt from Englishmen to pervert prosody. However, in tiie next breath, he makes an admission which upsets Mr. Mitford's theory of double accent ; ' for sometimes,' he says,' the necessary rapidity of utterance gives no room at all for accent, till we reach the fourth, fifth, or sixth syllable.' Now, in verse, as in music, acceleration of time never stifles accent ; therefore, the rapidity of utterance, here spoken of, can be due to no cause save the necessity of compressing within one bar (that is, in the intei-val be- tween two beats) three, four, or five successive weak syllables ; but if a word of four syllables, like ' ritrovarsi,' or one of three, like ' servitu,' can receive double accent, so can aU other words of similar dimensions ; and a fortiori, therefore, all words of five, six, or seven syllables ; and vice versa. But Mr. Mitford shifts his ground, so that one knows not where to have him : at one time, he talks of words needing two or more accents for verse- purposes in certain positions ; at others, of verse practically not needing accent at all in those positions : ' its remission,' he says, ' is an indispensable variety, required in almost every line: even its absence may sometimes be allowed, or, however, the remission may be such as to leave the character of an acute accent but doubtfully perceptible to the ear ' (p. 109) ; and he com- plains of the ' rude untutored ears ' which, incapable of perceiving the har- mony of verse, under such conditions, would require a forced pronunciation of the words (p. 93). The complaint is just ; but quite misplaced in the mouth of any one who adopts the theory of five feet for each verse. Besides, it were to be wished that Mr. Mitford had explained wherein consists the difference between a remission of accent and its absence. What sort of thing is a remitted stress ? and how ought it to be rendered ? PAUSES, AND THE PHRASING OF VERSE. 31 not cease) to be accented ; but that, being by position sub- ordinate, it ceases to count for accented in its bearing on the rhythmical movement.* And here it may be well to treat of pauses, and the man- ner of phrasing verse. All rhythm is divided into parts, equal as to duration of time, and unequal as to number of notes or syllables : the beats of verse-time depend on accented syllables in certain positions : syllables included between beats of time range themselves in rhythmical phrases ; and between each phrase and the next there is generally in our language a pause, slight indeed, yet quite perceptible, if tested by the more rapid transition of voice which takes place between syllables inseparably united in one and the same phrase. The manner of some, in phrasing verse, is so to prolong accent, or so to suspend the voice after the effort of accentua- tion, that every syllable which may follow, being part of the accented word, or in grammatical connexion with it, is thrown into a separate phrase ; thus, Where smi|ling spring j its ear|liest vi|sitspaid— : but this method, though quite admissible, and of good effect, if used with discretion, is sure to displease if habitually used : thus, if we were to continue it through the next line — , And part|ing sum|mer's ling'jring bloom | delay'd — , we should find ourselves lapsing into a sort of sing-song. The better way, in my opinion, is to treat as a phrase, * Under the following conditions, and perhaps under more, there exists between concurrent accented syllables a union so close as to admit of no severance : — 1. Between adjectives and participles, and the substantive they qualify : ex. 'cold ground,' ' bright pavement,' * serene air :' 2. Between a noun-substantive nominative and its verb, and vice versa. : ex, ' the king spake,' 'replied Henry :' 3. Between two substantives connected by the letter ' s,' as sign of the genitive case : ex. ' the queen's army,' ' the man's house :' 4. Between a verb and its regime : ex. ' gave order,' ' pronounced good :' 5. Between monosyllabic adverbs and adjectives like ' first,' ' last," still,' and a verb following : ex. 'first sprung,' ' still lives,' and 'vice versS,: ' ex. ' spoke last,' ' came first :' 6. Between a dissyllabic preposition and a substantive and participle following : ex. ' without cause,' ' within hearing,' ' about twenty.' 32 PHRASING OF TERSE. complete in itself, each accented word, or group of words, which can properly be uttered in one bar of strict declama- tive time : * Where smiling | spring f | its earliest | visits | paid — , And heap'd | the camp | with mountains | of the slain — , And dire | | imagination | still | pursues me — , And spirit | of the nethermost | abyss. But even this method, though it never results in sing- song, is apt to prove wearisome, if carried out to the strict letter. Most lines, then, may be phrased in more ways than one ; and there are few which may not be phrased in several ways ; for it is often impossible to say that the weak syllables which come between any two beats form necessarily part of one phrase more than of another. In fact, if the accents are pro- perly placed, the unaccented parts may be left to range them- selves at discretion. According, then, to the mode of phrasing adopted in any verse, or section of a verse, will be the position of pauses : if we adopt the first-mentioned mode, the pause will be on, or after, each accented syllable ; if the other, it will be at the close of each accented word. * The characteristics of declamation, as compared with common speech, have already been stated in p. 23. As to the length of time during which accented syllables should be held, no rule can be given : all we can say with contidence on this point, is, that the greater or less duration possible will be in proportion to the number of syllables included within each phrase. f Whether in those cases where a phrase consists of a single monosyl- lable, the time is equalised by dwelling longer on the syllable, or by a longer pause, or partly by one mode, partly by the other, I do not pretend to de- cide : nor is it a point of the least practical importance. All we need to be assured of is, that by some natural process the ear and voice do accommodate themselves to a great inequality of syllables in different phrases. X Sheridan says that in prose, adjectives are necessarily united to the substantive they qualify. If this be so, they are not less united in verse ; for whatever is good for prose is good, ' caeteris paribus,' for verse. But the fact is not as he states it ; except when the law of accentual subordination comes into play, that is, where the final syllable of the adjective and the initial syllable of the substantive are both accented, as in ' sweet Auburn,' NJold climate ; ' when this happens, there cannot be a severance ; but under other conditions, as in Departure from this happy place, our sweet Recess, and only consolation left — , ^a pause takes place just as naturally between the adjective and substantive as between any other two accented words. PHRASING OF VERSE. 33 The best manner of recitation is that which judiciously mixes the two modes. Generally, when a phrase consists of more than one word, the group will be found to have in it not more than one accent ; but there are exceptions ; for besides those cases in which, by the law above mentioned, one accented syllable becomes subordinate to another, it happens also that where the accentuation proceeds by thirds, as in Justly hast | in derision, and secure — , Sanguine, such | as celestial spirits bleed — , Present ? Thus | to his son audibly spake—, two accented syllables (neither of them being subordinate) will be uttered to one beat of time, even though there be a sentential stop between them; but this is not an effect peculiar to poetry ; for it exists, like all legitimate rhythmi- cal effects, as much in measured prose, as in verse. So that the phrasing of measured speech, on which pauses depend, is ruled partly by grammatical affinities, partly by prosodial ; partly by tendencies which cannot well be refen"ed either to grammar or prosody.* • The term Caesura, as it regards modern languages, is generally applied to those divisions of verse which are caused by sentential stops, thus : Of man's first disobedience, | and the fruit Of that forbidden tree|, whose mortal taste Brouerht death into the world], and all our woe — ; and in this sense I understand the term : it is also applied to an imaginary interval, which, in the absence of any stop, is supposed to divide verse, at some point, into two or more parts. Now, stops may coincide with pauses ; but they do not necessarily coincide ; for in many a line there is no stop ; whereas there are few English heroic lines without at least three pauses. The pause intervenes between one phrase and another ; but a stop may occur (as is shown above) between mem bers of one and the same phrase; and when this happens, the time given to the words is necessarily lessened by the time given to the stop ; though the latter, indeed, may be reduced to almost nothing ; for experience shows that not even by full stops ueed the rhythmical movement be interrupted ; and if, p recitation, the voice rests, so as to retard the time, on account of them, this is an effect of utterance de- pendent wholly on the taste and judgment of the reciter. And if it be said that, according to this theory, there will be no difference whether the sense ends with each verse, or whether it be ' variously drawn out from one verse to another,' I reply that the ear, being an intelligent organ, never fails to take account of these sentential divisions, even though they be not marked by any distinctive suspension of vocal sound. Now, the reason why sentential stops have no distinctive effect on verse is, I think, this : in common speech, the transition from word to word of each sentence is not less rapid than that D 34 SHERIDAN ON MILTON. Regarding pauses, then, as natural effects of language, and acknowledging, for rhythmical purposes, no difference of kind between them, I am quite at variance with Sheridan when he says that if the first thirteen lines of * Paradise Lost ' were written as prose, and read by some one who had never seen the poem, they would not easily be taken for verse : * on the contrary, I hold that these lines will well bear such a from syllable to syllable of each word ; but between sentences, stops are, for the most part, observed : now, recited verse is declamation, the time of which is slow enough for phrases to be distinguished from each other ; and as in declaimed speech phrases are divided by a process similar to that whereby sentences are divided in common speech, it must follow that unless stops be lengthened in the former proportionately to the slowness of declama- tion time, there will be no appreciable difference between sentential stops and rhythmical pauses. Liiidley Murray asserts that this proportionate prolongation takes place : I assert that it does not, and cannot, without causing a series of misplaced vacuities. As regards the form of caesura which is attributed in verse where there is no stop, as in From off the tossing of these fiery waves — , Sonorous metal blowing martial sound — , I do not see what reason there is for assuming a pause to exist between any two given words more than between any other given two. * Sheridan's views (which are echoed by Lindley Murray), on this point, will be found set forth from p. 102 to p. 116 in the second part of his ' Art of Reading.' Nowhere in the language are to be found better lines than the thirteen above mentioned ; and if such as they sound prosaic, or worse, when pro- nounced as prose, the same thing may be said of all poetry whatever. But if poets themselves submit their numbers to be thus written and pronounced, we may presume them to understand what they are about. VVell, just in this way did Alfieri treat his own tragic verse, with a view to secure it against the sing-song pronunciation of ill-taught actresses. 'Un ottimo secreto,' he says, ' per farle recitare a senso, e non cantare a verso a verso, come sogliono, sara di dar loro la parte scritta come fosse prosa.' (' Par- ere sull' arte comica.') Now, the verse thus treated was wrought with extreme care : it is not after the model of Petrarch's verse, nor of Ariosto's, nor of Tasso's, still less of Metastasio's ; but it is strictly coxTect ; and pre- serves, throughout, the movement which the author deemed best suited to the severe grandeur of tragedy : and 3'et, for all that, it dwindles, forsooth, to prose in recitation, unless helped out by an artificial utterance ! But the thing Alfieri most shunned was this very artificialness : written out as prose, his verse was by himself placed in the hands of actors, with ex- pi-ess purpose that it should have, in their mouths, the proper (that is, the natural) pronunciation of prose. And what made him so confident ? Per- ception of a simple truth, that verse, pronounced properly, takes care of itself, i properly constructed. If Sheridan's remarks, bearing directly and indirectly on poetic pauses, be brought together and summarised, they are found to be just as confused and self-contradictory as his remarks on quantity and accent. For instance : All words, he tells us, should be pronounced in poetry as they are pronounced SHERIDAN ON MILTON. 35 test ; and that no combination of words which will not bear it deserves to be called verse. in prose ; ' and yet, to read poetry, and to read prose, are two quite different things ; * for if the former be read as though it were prose, that is, without some treatment, peculiar to verse,^ as regards pauses, there is no possibility of distinguishing one from the other ; '^ nay, verse is turned into poetical prose, or rather into prose run mad : ^ there are, then, two kinds of reading, one suitable to prose, one to poetry ; and there are two kinds of pauses, utterly distinct from each other,^ namely, the ' sentential ' and the ' poetic : ' the office of the one being to point out the sense, that of the other to mark the melody : ^ ' sentential ' pauses are those which are known as ' stops,' 8 and they are always marked by a change of note ; ^ and in this change of note consists, it seems, the disjunctive element of the pause : lo of the poetic kind there are two, the 'ciesural' and the ' final,' ^i exactly similar in character and governed by the same laws,^^ i^ut differing in position ; and the difference is, that the ' caesural ' occurs in the body of a verse, the 'final' at its close: i^ these pauses sometimes coincide with the sentential (notwithstanding the utter distinctness of the two kinds), some- times have an independent state ; ^^ that is, exist where there is no pause in the sense ; and they consist only in a suspension of the voice,'-^ without that change of note which sentential pauses always, as is said, have. Now, certain Avords are necessarily connected in sentences ; ^^ go that to separate them by any sort of pause in prose is contrary to the genius of our language ; ^^ and so it would be in poetry if the effect of the pauses were here the same ; but the fact is, that when such words occur, one at the end of a verse, another at the beginning of the next verse, they can be separated by the * final ' pause, so as to mark the melody, without confounding the sense ; ^** for as the sentential stops, which affect the sense, alone possess, in change of note, the true element of disjunction, the sense cannot in verse (though it may in prose i^) be affected bj^ any pause which does not carry with it the said change of note : '^'^ hence a final pause, which, after all, is but a suspension of the voice, does not really separate the words kept apart by it while the suspension lasts ; on the contrary, they remain, with the pause between theni, every bit as well united as if they had been pronounced con- junctively .21 But though this law is good as regards the operation of the ' final ' pause, it does not hold good as regards the caesural (which, however, is of a character exactly similar, and governed by the same laws) ; for, in order to find out the seat of the ' caesura,' we must reflect that there are, indeed, some parts of speech so necessarily connected in sentences as not to admit of any separation by slightest pause of the voice ; and between such words, therefore, the caesura never falls ; ^''^ at least ' never,' except after a word which leaves an idea for the mind to dwell on ;23 but this exception may pro- perly be made, as, for instance, in Aml)ition first" sprang from your blest abodes — ; for though all words, other than particles, leave some idea on the mind, and though in prose there would not properly be here a pause after * first,' on account of its close connexion with the following words, yet in verse there ' Vol. i. p. 292. " Ibid. pp. 103, 104, 105, 108, 111, 112, 118, 257. ' Ibid. pp.103, 107, 108,293. * Ibid. pp. 103, 104, 113, 116. ^ Ibid. pp. 104, 105,111. « Ibid. p. 107. ' Ibid. pp. 107, 117. « Ibid. p. 107. « Ibid, pp. 112, 113. '" Ibid. pp. 117, 257, 258. " Ibid. pp. 102, 293. " Ibid. p. 117. " Ibid, p. 102. " Ibid. p. 117. " Ibid. pp. 112, 113, 114. ^° Ibid. pp. 109, 257. 258, 294, 298. " Ibid. p. 257. '« Ibid. pp. 107, 112, 113. '" Ibid. pp. 257, 294. ^» Ibid. pp. 112, 113. " Ibid. ii. p. 258. ="=■ Ibid. pp. 293, 294. " Ibid. p. 294. d2 36 CONCURRENT ACCENTS. Now, it was made to be understood above, that the inter- vention of a pause may nullify the operation of the afore- said law relating to concurrent accents : this is seen in the verses ah-eady quoted : Amaz'd I stood | harass 'd with grief and fear — , Ease to the body some, | none to the mind — : in such cases the syllables have no connexion with each other ; a pause takes place between them ; and each receives metrical accent. But though such concurrent accented syllables as naturally unite must not be severed, that is, must not form parts of separate phrases, it does not follow that disconnected syllables may not form part of one and the same phrase. In the verse following — , Kocks, caves, lakes, fens, bogs, dens, and shades of death — , there are six consecutive accented syllables, no one of them being connected with another, and no one, more than another, disconnected ; and yet, however distinctly each be uttered, they are found to range themselves, two by two, into rhythmical phrases, in which the first, the third, and the fifth syllables are respectively subordinate to the second, the fourth, and the sixth. If there be a series of seven such consecutive accented syllables, the seventh will take metrical accent; and the word, whether it be a m.onosyllable or a dissyllable, will form a phrase by itself. When the first of any two such accented syllables (that must be one ; because, if you remove the caesura, the metre is entirely de- stroyed : 1 at the same time, we should take care never to place the caesura after the adjective in a line like Your own resistless" eloquence employ — ; because such an unnatural disjunction of words which have a necessary con- nexion with each other, whatever pleasure it may give to the ear, must hurt the understanding, which surely, in rational beings, has the best claim to be satisfied. ^ As regards his assertion that sentential stops are marked by change of note, I simply deny the fact so to be. » Vol. u. p. 294. » Ibid. pp. 297, 298. THE FIFTH SYLLABLE. 37 is, any two not necessarily united), is an even one (namely, the second, fourth, sixth, or eighth), as in 2nd Inclines | here to continue, and build up here — , 4th And out of good, [ still to find means of evil — , 6:h Far round illumin'd hell ; [ highly they rage—, the accent of such syllables is not affected by that of any other syllable immediately following; but both will have metrical accent : if, on the other hand, the first of the pair be an uneven one, as happens in And in good still | to find the means of evil — , And on wing pois'd | beneath the cope of heav'n — , they must be united in one phrase ; for heroic verse does not admit of metrical accent on uneven syllables, except under conditions which do not exist here ; and the syllables in ques- tion may be so phrased, or not, at pleasure ; for though there is, indeed, no necessary cohesion between them, yet neither is there any necessary disjunction.* The fifth syllable must never be metrically accented unless the fourth be so, and be followed by a pause, as in 4th 5th And casts a gloom over this tufted grove — : in the lines following. High eminent, blooming ambrosial fruit — , « The vanquisher spoil'd of his boasted spoil — , the fourth, though followed by a pause, is not accented ; and the rhjiihm is, therefore, faulty. * Thus, on the same principle, if the measure were anapaestic, they might be phrased separately : And la good [still may find] many means] to work evil — , And on wing | pois'd beneath] the blue cope | of the skies — : or again, conjointly : In good still I may find means], many means], to work evil — , On wing pois'd] underneath] the blue cope]of the skies — : which ever arrangement we elect, each seems as natural as the- other. It may, then, be taken as a general law, that where concurrent accented syllables naturally cohere, they should, on no account, be disjoined to suit the exigencies of metre ; those, on the other hand, which have no such prin- ciple of cohesion, may be phrased together, or separately, just as the metre requires. 38 THE FIFTH SYLLABLE. This next, 4th 5th Who with his soft pipe and smooth dittied lay — ,* is faulty likewise ; because here the fourth syllable, though accented, is not metrically accented, being subordinate to the fifth ; and subordinate it must be, since there is no possibility of a pause between them. If, then, the fifth syllable be accented, under conditions other than the one above stated, there is an absolute necessity that the sixth should be accented also, so as to resolve the dissonance created by the accentuation of the fifth : thus, 5th Gth Alone, and without guide, half lost I seek — , 5th 6th Undaunted to seek there whatever foe — .f The seventh syllable (as we have seen) may be metrically accented, if either the accentual movement be from the fourth to the seventh and tenth, or if the sixth be accented and followed by a pause ; but, failing thes3 conditions, the seventh should not be accented, unless the eighth be also, as in 7th «th Mov'd on in silence to soft pipes that charm'd — : hence the line Thy ling'ring, or with one stroke of this dart — , | is faulty in the sixth and seventh syllables, for the same reason as the above-quoted line. * Lines of this form are, however, very common in English poetry of the past and present time. t Except under the conditions specified, heroic verse abhors metrical accent on the fifth syllable : and yet lines, having this peculiarity, though hopelessly lame considered as heroic verse, may be free from fault, considered as anapaestic verse of ten syllables. Such is the case as regards the lines My vanquisher spoil'd of his boasted spoil — , High eminent, blooming ambrosial fruit — , and many others. Still, it does not follow that heroic verse rejects all anapaestic forms. When accented on the third and sixth syllables, or on the seventh and tenth, it has an anapaestic movement in the first or last hemistich ; and is itself an anapaestic verse of ten syllables. The truth then is, that both these kinds of verse borrow something from each other, but not everything ; some forms, but not all forms : thus heroic verse accepts those anapaestic forms, and those only, which allow it to have metrical accent on the fourth or sixth syllables ; in other words, it accepts such forms in the first hemistich or the second, but never in both: now, the lines above quoted have accent on neither the fourth nor sixth syllable ; and are ana- paestic in each hemistich. X Sheridan strives to make out that this line, if properly read, is not lame. ' If the verse,' he says, *be pronounced thus THE SEYENTH SYLLABLE. 39 Who with his soft pipe, etc., is faulty in the fifth. The verse In curls on either cheek play'd : wings he wore—, does not to me seem defective ; but there is a peculiarity about it which needs examination. The sixth, seventh, and eighth syllables are accented ; and between the seventh and eighth there is a colon, if not a full stop. The phrasing, I think, should be as follows — , 6th Tth 8th In curls I on either | cheek play'd : | wings | he wore | — : thus the sixth and eighth syllables will have metrical accent, and the seventh will be subordinate to the eighth ; still, the sixth and seventh ought not to be separated ; for though there is no necessary cohesion between an oblique case of a noun-substantive and a transitive vei'b following, yet there seems to be such between an oblique case and an intransitive verb, following, and closing the sentence. The seventh syllable, then, will be subordinate to the eighth, notwithstanding that they form part of separate Thy ling' ring, or with bne stro'ke of this dart, the verse will be degraded to hobbling prose. And although it may be im- agined that the sense is preserved in this way of reading, yet it will appear, on examination, that part of the poet's meaning is lost, as well as the imagery, to preserve which there must be a strong emphasis on the words "one" and "this," as thus — : Thy ling'ring, or with on'e stroke of thi's dart : for the emphasis on the word " one ' ' marks the pecidiar property of the dart of Death, which does its business at once, and needs no second stroke ; and that on the word " this" presents the dart to view, and the image of death shaking it at Satan.' (Vol. ii. pp. 279, 280.) The words ' one ' and ' this ' might here, I should say, have been left alone ; seeing that on the face of them they carry their full meaning : no numeral counts for more than its own number, and nothing signified by a demonstrative pronoun can be otherwise than presented to ocular or mental view. Besides, what we want to mend in this line is not emphasis, but a severance of the sixth syllable from the seventh ; and if prosodial accent cannot do it, neither could emphasis, if there were any ; but there is none ; for as nobody asserts that two strokes would be needed, it is futile to insist, in impressive tones, that they Avould not. Again 'this' is not em- phasised, for it is not even accented ; and it could not be accented unless there were antithesis, and there is none here ; for, assuredly, no question is raised as between the particular dart specified and any other. 40 THE THIRD SYLLABLE. phrases, and that there is a sentential full stop between them ; in other words, the accent on the seventh syllable creates a dissonance which awaits resolution; and as the -resolution takes place on the eighth, the ear is satisfied. Also accented on the sixth, seventh, and eighth syllables are the verses following, 6th 7th 8th Threw forth, till from the left side op'ning wide — , 6th 7th 8th Approach not, but with both wings veil their eyes — : here, also, there can be no severance between the sixth and seventh ; and the seventh is subordinate to the eighth. In these next, 2nd 3rd 4th Where all life dies, death lives, and nature breeds — , 2nd 3rd 4th Our own right lostj Him to unthrone we then — , there is, on the same principle, no severance between the sscond and third, and the third is subordinate to the fourth. In short, whether the accentuation be on the second, third, and fourth; the fourth, fifth, and sixth; the sixth, seventh, and eighth ; or the eighth, ninth, and tenth ; similar effects will always be found to present themselves, under similar conditions.* * The followins: passage from a well-known author may be here read with advantage : ' As the proper mixture of light and shade has a noble effect in painting, so a judicious mixture of concords and discords is equally essential to a musical composition : as shades are necessary to relieve the eye, which is soon tired and disgusted with a level glare of light, so dis- cords are necessary to relieve the ear, which is otherwise immediately satiated with a continued and unvaried strain of harmony. We may add (for the sake of those who are not in any degree acquainted with the theory of music) that the Preparations and Resolutions of discords resemble the soft gradations from light to shade, or from shade to light, in Painting.' ('Essay on Musical Expression.' By Charles Avison.) When non-metrical accents are so placed as to offend no law of verse, they may fitly be likened to these harmonic discords which, causing variety l?y force of contrast and suspended satisfaction, are themselves to music (as Avison well says) what shade is to landscape. But with us there prevails a notion that the right way to ensure this sort of variety in verse is to alternate between law and Ucense ; to gratify the ear and disgust it by turns ; to give first so much pleasure, and then, by way of foil, so much pain ; after the pain, then again, by way of compensation, so much plea- sure : and those who take this view are evermore pointing to the supposed analogy of musical discords ; as though the effect of them were like that of SIMILAK CONDITIONS, SIMILAR EFFECTS. 41 But not only in the internal structure of the verse, at the close of it also, and at the beginning of the next one, must account be taken both of pauses, and of the action on each other of concurrent accents : for in our decasyllabic metre there is often a danger lest when the second of any two verses begins with an accented syllable the operation of the above-mentioned law should spoil the rhythm. This will not happen, indeed, where the arrangement is such as to permit of a pause between the final and initial syllables; but it does happen whenever they have a relation to each other so close as to render a pause impossible, as in the verses following, — 1. Went up, and water'd all the ground, and each Plant of the field — , 2. His longitude throuofh heav'n's high road ; the grey Dawn, and the Pleiades before him danc'd — , 3. One foot he center'd, and the other turn'd Bound, through the vast profundity obscure — , 4. Defaming as impure, what God declares Pure,* and commands to some, leaves free to all — : ill-tuned instruments, or hap-hazard smashes of sound, and must needs therefore have some offence : whereas the fact is far otherwise ; for not less subject to strict rule, nor less tending to pleasurable effect, is the use of dis- cords in music than the use of concords, or aught else in that delicate and complex science. The true theory of variety in composition is well stated by Dionysius : * I have, thirdly,' he says, ' to speak of change, as among those causes which make harmony beautiful : but 1 mean not a change from better to worse (for that would be sheer folly), nor yet one from worse to better, but variety in things of like nature : for all things beautiful, and all that are sweet, bring satiety if kept in one and the same state ; but, if varied with change, they keep new for ever.' (Sec. XIX.) And here I will say, once for all, that within the strict limits of legiti- mate versification, as laid down in this treatise, there is ample scope for the utmost variety verse can ever need. * With reference to these last quoted lines. Dr. Johnson says, * When a syllable is cut off from the rest, it must either be united to the line with which the sense connects it, or be sounded alone. If it be united to the other line, it corrupts the harmony : if disjoined, it must stand alone, and, with regard to music, be superfluous : for there is no harmony in a single sound, because it has no proportion to another.' (Rambler, No. 90.) Agreeing with Dr. Johnson that if the syllables in question be united, they corrupt the harmony, I do not agree that if they be disjoined the second syllable is superfluous or unmusical, as regards the second line : for though, doubtless, there is no harmony in a single sound, yet in considering any series of sounds arranged for musical effect we have no right to take one 42 FINAL PAUSE. now here, between ' each ' and ' plant/ between ' grey ' and * dawn/ between ' turn'd ' and ' round/ between * de- clares ' and ' pure/ there cannot possibly be any severance : nevertheless, this impossible thing is in each case needed to save the versification from manifest faultiness. In the lines, And bark with frizzled hair implicit. Last Rose, as in dance, the stately trees — , * last ' and ' rose ' may be disjoined, or not, according as the monosyllable ' last ' be taken as an adverb, signifying * finally,' or ' in conclusion,' or whether it be taken as an adjective, signifying, as regards the ti-ees, the quality of being the last to rise : in the former case, there would be a pause between * last ' and ' rose ; ' in the latter none ; so that the verse would be faulty, or not, according to the sense we elect. I do not know that anything more remains for me to say touching heroic decasyllabic verse ; but with reference to the opinion already given that some forms of this metre do not well adapt themselves to our rhyming heroic couplet, I will sound apart from the rest, and say there is no music in it, but we are bound to consider the whole series together, and on that pass judgment : thus tried, and taken by itself, the verse, Pure, and commands to some, leaves free to all, is really faultless : but if it be taken in connexion with the preceding one, both must be pronounced faulty ; because only by falsifying the pronun- ciation of the phrase 'declares pure' (which is divided between them) can the monosyllable ' pure ' be made to sound as belonging to the verse in which it claims a place. And that to this cause, and not to the one assigned by Dr. Johnson, the faultiness is due, may, I think, be shown by the follow- ing verses — , Seasons return : but not to me returns Day, or the sweet approach of ev'n or morn — ; for here the hemistich, ' Day or the sweet approach ' is exactly similar to the hemistich, ' Pure and commands to some ; ' so that if the latter be de- fective, so must also be the former ; and if it be a thing essentially objec- tionable for the second of two given lines to begin with an accented monosyllable which the sense connects with a preceding decasyllabic line, this objection must apply to both cases equallj'- : but the hemistich ' Day or the sweet approach ' is not found to be defective in itself ; nor is the ear offended by the transition in these lines, from ' returns ' to ' day,' as it is in the others, by the transition from * declares ' to ' pure. ' Why not then in both cases if in one ? Because in one of them the words are too closely united to admit of an intervening pause, while in the other they are not so closely united. OTHER IAMBIC METRES. 43 here add that of the thirty-five forms, heretofore mentioned, those which have the best effect in it are the 1st, 2nd, 8th, 9th, 12th, 14th, 16thj 22nd, and 29th ; and that the remain- ing twenty-six, while suitable, more or less, for the spense- rean stanza, rima ottava or terza, and blank verse, do very seldom, if ever, find a place in the polished couplets of Pope, Johnson,* Goldsmith, and Byron, or those of the Rolliad and the Anti- Jacobin. Before quitting this part of my subject, I will here say a few words about other iambic metres. Alexandine verse has twelve syllables, the twelfth being accented ; and its primary form consists of six accentual iambi, as the heroic consists of five : e.g.. The which to hear vouchsafe, dearest friend, awhile :t we have also an iambic metre, consisting primarily of fourteen syllables, and of seven iambi : e.g.. The townsmen shrank from right and left, and ey'd askance witly • «| fear 7 %J His low'ring brow, and curling lip, which always seem'd to sneer : 1 ^v^ both these kinds of verse admit a variable number of ac- cents; though not so many as heroic verse admits, nor so * Dr. Johnson's * Vanity of Human Wishes ' is scarcely inferior, in point of versification, to any heroic couplet poem of our language ; and it is always read with pleasure : his tragedy, ' Irene,' is not less well versified, so far as mere absence of fault goes ; but we can scarce endure to read two pages of it: and although this may be due to more causes than one, yet monotony of rhythm is, beyond doubt, one chief cause. Now, the rhyth- mical movement of both these poems is the same ; in both it ranges within the same narrow groove ; and one of them seems monotonous while the other does not, because our heroic couplet is then found to be most effective when it does not exceed the few varieties of form above mentioned ; where- as blank verse not only admits, but requires, a far greater variety. t Professor Craik says ('Verse of Shakespear,' ]». 35) that the character- istic of an alexandrine consists in pressure on the sixth syllable and on the twelfth : but there is really no such law : the pressure mav happen to be more often on the sixtli than on the fourth ; yet if it be on 'the fourth, and not on the sixth, the verse does not suffer. The following lines of Spenser afford examples of such an arrangement : Their minds to pleasure, and their mouths to dainty fare—. That made him reel, and to his breast his beaver bent — , Who, nought regarding his displeasure, forward go'th— ; and I fail to see that they lack anything necessary. 44 GENEKAL LAWS. many variations in other respects. Neither of them ap- pears, however, to have been used in the Romance languages.* Touching the smaller iambic metres, all I have to say is that verse of eight syllables may have four, three, or two metrical accents, and verse of six syllables, three, two, or, possibly, but one. Recurring now to the general laws of verse, I will con- sider them separately in their application to the poetry of our language. As moflern verse, then, requires accent in certain posi- tions, and as no word has more than one accent, while a great many have none at all, it follows that whenever in those parts of a verse which require metrical accent there occurs a syllable not entitled to bear accent, the verse will be faulty. Now, this fault, which is scrupulously avoided by poets of the Romance languages, is of frequent Occurrence in English poetry ; as will at once be manifest if, passing by other cases, we do but consider how great is the number of English heroic verses having for their tenth syllable, which invariably needs accent, either the final syllable of a word like * villager,' * harmony,' * astonishment ' (which, beyond doubt, are accented, not on their final syllable, but on their antepenultimate), or the personal pronoun, under conditions which do not allow of its being accented, or some insignificant proclytic f particle, like ' to,' ' for,' ' of,' * the,' ' and,' which under no circumstance whatever can bear accent. * That is, other than the French. French heroic verse, indeed, is said to be alexandrine : but its claim to be so considered rests wholly on the assump- tion that feminine (that is, mute) endings are something more than mere ocular effects ; which, except in song, they do not seem to be. As regards mute endings, see what says the Abbe d'Olivet in his * Prosodie fran^aise.' The treatise is now rare ; but it is to be found in- serted, by way of preface, in some quarto editions of Boyer's Dictionary. f By the term * proclytic,' I mean any part of speech which leans pre- positively on an accented word. In the phrase, * but that it may be seen,' there are no fewer than five proclytics, all leaning on ' seen.' Even a dissyllable may be proclytic : such, at least, is ' upon; ' which always, when preceding a noun, attaches itself inseparably thereto, and loses its own accent : following a verb, as in ' falls upon,' it becomes encly- tic, that is, inseparably attached to an antecedent M-ord ; and here also it ceases to be accented. ENGLISH LICENSE. 45 Verses ending with a weak tenth syllable we meet at every turn in English poetry ; and I doubt if (excepting Goldsmith) there be one among our poets whose works are free from this disfigurement. But the worst form of the fault is unquestionably that in which insignificant particles are made to serve ; because, in such cases, besides the misuse of a weak syllable where a strong one is needed, there is the further inconvenience of separating two things by nature meant to be inseparable.* This form, little countenanced by Milton, and quite re- jected by many since his time, is very much in vogue with English dramatists. The examples following are from Shakespear and Massinger : 1. Come, my sweet wife, my dearest mother, and My friends of noble touch — , 2. What iprosperously I have attempted and With bloody passage led your arms ev'n to The walls of Rome — , S. 1. Of sorrow : she appear'd the mistress of Most rare perfections — , 2. The ceremony which he uses in Bestowing — , 3. Then he's my husband's son, the fitter to Supply my wants — ; M. and these next are from Lord Byron : 1. Then thou must doff the ducal coronet from That head—, 2. Ne'er rear'd their sanguinary shadows to Confront a traitor — , 3. I cannot plead to my inferior, nor Can recognise your legal pow'r — , 4. Your sin hath made us make a law which will Become a precedent — , * Except when standing alone, signs of comparison and demonstrative pronouns should not be used at the close of a verse : because, though often accented, they do not well admit after them the pause which every verse needs in that position. Still, they may be used, to point metre, elsewhere ; for an interval of time does not necessarily occur between each phrase and the next in verse, any more tlian between each note and the next in music. And when I speak of intervals, I speak but of such as the ear easily appre- ciates : now such a one, though needed between verse and verse, is not needed between phrase and phrase. (See p. 12.) 46 OCULAR SYLLABLES. 5. And who be they ? In numbers many, but The first now stands before you — , 6. Not ev'n contented with a sceptre, till They can convert it to a two-edged sword — , 7. You singled me out like a victim to Stand crown'd — , 8. You had ev'n in the interregnum of My journey to the Capitol — , 9. Yet left the Duke. All this I bore, and would Have borne, until — , 10. Is here in virtue of his office, as One of the forty — . * The third rule, which is, that no syllable, not sounded in measured prose, should be sounded in verse, we find carefully observed by Pope, and the best among his followers : but by Milton and our dramatists it is set at naught without scruple. This happens in three ways : firstly, when two vowels which make, to the ear, but one syllable, as, for instance, in * delu- sion,' * notion,' ' partial,' * ocean,' * glorious,' ' courtier,' ' mar- riage,' ' conscience,' * clear,' * your,' are dissolved, that is, separated into two distinct sounds, f as is the case in the verses following : — How bitter is such self -delusion ! J — , * Tried by the proportion which lame lines bear to sound, there is no- thing in the language lamer than Lord Byron's dramatic blank-verse. The above-given ten examples of one fault all occur within the compass of seventy-five lines in the fifth act of 'Marino Faliero,' and there are numerous faults of other kinds within the same compass. Perhaps it would not be easy to show elsewhere, in his lordship's dramas, seventy-five consecutive lines quite so bad as these : nevertheless, the general character of his dra- matic verse is such as fully to justify the critic who describes it as ' often only distinguishable from prose by the unrelenting uniformity with which it is divided into decasyllabic portions.' (Reginald Heber, ' Quart. Rev.' vol. xxvii.) t The practice is very common in Spenser's ' Fairy Queen,' and in all the dramas of the Elizabethan era. J Lines of this kind offend simultaneously against the second and third rule ; and the following, — Where with her best nurse Contemplation — , has the additional fault of a dissonance unresolved on the fifth syllable. But dissolution of the final, ' ion,' though frequent in Comus, disappears in Mil- ton's later works. As the endings of words like ' delusion,' * nation,' ' partial,' ' ocean' (to take these as examples), are really 'zhon,' 'shon,' 'shall,' and 'shan,'the effect of dissolution in such cases, is not to separate a diplithong (for there is none) into its elements, but to create a syllabic sound out of nothing. ELONaATED SYLLABLES. 47 With all the grisly legions that troop — , And so all yoiirs. these naughty times — , And so though yours, not yoiirs. Prove it so — , Dear my lord, if you, in your reproof — : secondly, when, by a cognate process, words like ' fire,' ' de- sire,' * sure,' 'kindred,' * England,' ' angry,' ' children,' ' content- ment,' are artificially lengthened ; — ' fire ' and ' sure ' becoming dissyllables, * desire ' and ' England' trisyllables, and so on:* or Gilford, alluding to the frequency of this practice in Massinger's verse, quaintly remarks : 'It may slightly embarra.>-s the reader at first, but a little acquaintance will show its advantages, and render il not only easy but delightful.' (Introduction, p. xxx.) Eels are said not much to mind being skinned alive when used to it ; but I never yet heard that use makes them like it. In French, there is a phrase about skinning people's ears ; and some people, outdoing the eels, may be brought, it seems, quite to enjoy such treatment. '* By way of accounting for apparent irregularities in Shakespear's dra- matic verse, Mr. VVm. Sydney Walker is at pains to show that our early dramatists habitually used these forms of elongation and forms of curtail- ment not less remarkable. But to account for and to justify are different things. Lines which halt, through lack or through excess of natural parts, are not cured by unduly sounding fictitious syllables, or by stifling real ones which ought to be sounded : such remedies are worse than the disease ; for though false metre be bad enough, it is not so bad as falsified language. The only way to justify sucii modes is to show that the pronunciation of English, at the time these authors wrote, was materially different from what it is now : but there is no reason to think it was : Ben Jonson's chap- ters on prosody might pass for the work of a modern grammatist ; and Shakespear's rule of right pronunciation we learn from his ' sonnets' and other * poems,' which scarce contain two words requiring a treatment different from that given them in the poetry of our time. Ben Jonson, it is true, says that all simple dissyllabic nouns (among which he instances 'belief ') are accented on their first syllable; and if the fact were so, it would prove a change to have taken place in the pronuncia- tion of many words since his time : but the fact is otherwise ; for there is hardly a word of this class which, accented now on its final syllable, may not be shown to have had the same accentuation three centuries ago ; and that ' belief does not form an exception is shown by the following extracts from Shakespear, — Stands not within the prospect of belief — , Which are to my belief, witness'd the rather — , ^ And let belief and life encounter so — , His Highness yet doth speak, and hold belief — , And will not let belief take hold of him — . Throughout the ' sonnets ' and ' poems ' of Shakespear, I observe but one instance of a word treated (from our point of view) exceptionally : it occurs thus, — If in thy hope thou dar'st do such outra'ge — , where ' outrage,' rhyming with ' age,' is made to bear accent on its final syllable ; but this is a clear case of poetical license, seeing that Shake- spear, in his dramas, always accents the word on its penultimate : e.g. In murder and in outrage bloody there — , I fear some outrage, and I'll follow thee — , 48 MUTE VOWELS. thii'dly, when syllabic sound is given to vowels mute between two consonants, as in words like ' misery,' ' interest,' * de- liverance,' ' severance,* ' temperate,' * conqueror,' * labouring,' * murmuring,' * flower,' ' heaven,' and numberless others : * but this form of the fault is, at times, found more par- donable than the before-mentioned ones ; because some of the vowels, though practically ignored in common speech, have still a latent syllabic reality. A few examples will suffice to show the opposite modes of treating some such words in verse : Spring from the venomous outrage of the Duke — , Do outrage, and displeasure to himself — . We may safely, then, I think, infer that the above-mentioned practice of our dramatists is not to be defended by pleading a difference between the pronunciation of their time and ours : indeed, we have direct evidence on the subject: for cotemporary authors were found protesting against the practice as an unwarrantable departure from the proper pronunciation of English. (See Guest's ' English Rhythms,' vol. i. pp. 182, 183.) * In Professor Craik's 'English of Shakespear,' I find the following commentary on the line. Being cross'd in conference with some senators : — 'If the "being" and the "conference" (he says) be fully enunciated, as they will be in any but the most slovenly reading, we have two supernu- merary syllables in this line, but both so short, that neither the mechanism, nor the melody, of the verse is at all impaired by them.' (Sec. 62, p. 109.) In face of what is here asserted touching the right of ' being ' to be held a dissyllable and ' conference ' a trisyllable, I have only to say that my own judgment leads me to the opposite conclusion. Moreover, if it be a most slovenly reading which does not so treat these words, most slov- enly also must be the versification which does not accredit them with the specified due proportions. Now, unquestionably, as seems to me, if the *i' of 'being,' and the intermediate 'e' of 'conference,' be separately sounded in the verse, the rhythm is spoilt. Professor Craik assumes that they are sounded ; only they are so short, he says, that the melody of the verse is not injured. This seems to me a contradiction : either they have that full enunciation without which, according to him, the reading is slov- enly, that is, either they are made separately perceptible to the ear, or not ; if they be, they cannot help impairing the rhythm : but the rhythm, we are told, is not impaired ; it follows, then, that the vowels are not separately sounded ; and such, in my judgment, is the proper treatment. As a test whether a bona fide utterance of these alleged syllables will spoil the verse, or not, let us suppose a line like the following, — Caesar spoke in confidence to some senators : — here the dissvllable 'Cfesar' stands in place of the dissyllable 'being,' and the trisyllable ' confidence ' in place of the trisyllable ' conference ; ' but because the final syllable of ' Caesar ' cannot be absorbed, nor the inter- mediate ' i ' of ' confidence ' be suppressed, as are, according to my view, the corresponding vowels of ' being,' and ' conference,' the result is a com- bination of words in which no one, I think, will detect the faintest semblance of metre. MONOSYLLABLES OR DISSYLLABLES? 49 The close of all my mis'ries, and the balm — , Began to parch the temp 'rate zone, whereat — , Pond 'ring the danger with deep thoughts, and each — , Yet ever plotting how the conq'ror least — : here, the words ' miseries/ * temperate/ * pondering/ * con- queror/ are rightly used as dissyllables ; * while in the examples next to come they are less well used as trisylla- bles, e.g. : — And prov'd the source of all my miseries — , Thy temperance, invincible besides — , So pondering, and from his arm^d side — , To adore the conqueror, who now beholds. Words like * tower,' * flower,' * heaven,' * seven,' * given,' * stolen,' * swollen,' have the manifest stamp of monosyllables ; nor are they ever treated otherwise than as such by the best vei-sifiers of the last century : similar, as a rule, is Milton's treatment of them; though, now and then, we find him making them do service as dissyllables. In the lines following they are rightly treated as mono- syllables : — Warriors, the flow'r of heav'n, once yours now lost — , His utmost pow'r with adverse pow'r oppos'd — , Obstruct heav'n 's tow'rs, and in derision sits — , See where it flows disgorging at sev'n mouths — , Till, as a signal giv'n, the uplifted spear — , If thou bee'st he — but oh 1 how fall'n, how chang'd — , Kis'n from the river o'er the moorish glades — , To whom the fiend, now swoll'n with rage, replied — : and here follow examples showing the contrary effect : — At which command, the powers * militant — , Had driven out the ungodly from his seat — , Is risen to invade us ; whom no less — . These words are monosyllabic, because, in the pronuncia- tion of them, we make but one movement of the vocal organs : thus, between the initial and final consonants of ' power,' there is the diphthong sound of ' ou ' or ' ow,' which, * Though words like * tower,' ' flower,' are of frequent occurrence in Milton, I am not able to put my hand on another example, besides thi^, iu which he uses any of them as dissyllables. E 50 WHY MONOSYLLABLES? combining with these consonants, makes one syllabic sound, the effect of which is exactly the same as that of similar com- binations in ' hour,' ' flour,' 'sour : ' * again, in the ordinary pronunciation of ' heaven,' ' driven,' ' risen,'! ' swollen,' there is no more than one movement of the lips and tongue ; for the sound of the final * en ' (which inaccurate observers are * In writing thus, I do not overlook the fact that there are some among us who confidently assert the proper pronunciation of these words to be dissyllabic. As bearing on this point, the following passage from a recent work on versification may here be noticed : — *The worst stumbling-block in English is the word "our," and the next, perhaps, the poetical " flower." The second of these is acknowledged to be of two syllables : the first, owing to our absurd mode of spelling, only one, though nowise different virtually from the other. "Flower," indeed, is more often than not in verse jotted down as " flow'r," with a mark of elision, and treated as a monosyllable : but in the minuteness so common among English short vowels unaccented, that is not sufficient, the vowel still remains as much as ever ; the mute required to pronounce the '* r " in " our," is no otherwise circumstanced. — Never mtist either, or any other resembling word, be treated as less than two syllables, under any circum- stances whatever. — Herein lies the awkwardness of " our " above all other words. Too insignificant to receive accent, it yet always demands the con- sideration of a dissyllable, which is rendered the more annoying by its frequent occurrence.' — (' A Practical Guide to Engl. Versification,' by E. Wadham, 1869, pp. 59, 60.) If ' our ' and ' flower ' be by their nature dissyllables, one does not see why there should be any awkwardness in giving elfect to a natural law : but since an awkwardness is felt, those who feel it would do well to observe what happens in the process of pronunciation. If the Hps remain (as they ought) open, there cannot possibly be more than one syllable ; if they close, two syllables are obtained ; and if then the effect annoys us, we may rest assured that this arises, not from the words themselves, but from something peculiar in our mode of treating them. f In words like * heaven,' ' seven,' * given,' ' driven,' ' risen,' the in- termediate consonant power attaches itself wholly to the prior vowel, which sounds as though it were followed by a double consonant, thus, ' hew,' 'sew,' 'givv,' 'rizz ;' but in words like 'frozen,' 'chosen,' 'graven,' ' haven,' that is, in words having their prior vowel pronounced with its pri- mary sound, the intermediate consonant attaches itself to the subsequent vowel, and makes with it a distinct syllable, as in That Shepherd who first taught the cho-sen seed — , , Pour'd never from her fro-zen loins, to pass — ; notwithstanding that Milton, in one of his finest passages, uses 'chosen' as a monosyllable, e. g., Immortal spir'ts, or have ye chos'n this spot ? The only exception to the rule is in the case of words like ' fallen,' ' stolen,' ' swollen,' where the vowel 'e' becomes mute between the letters '1' and 'n.' Words like ' often,' ' soften,' come within the rule ; for here the prior vowel sound is not ' o ' but ' au,' and the consonant ' t,' being mute, ' au ' and ' fen ' are sounded separately : even if the printed ' o ' be sounded curtly, as in ' scoff^,' the intermediate consonant power is found to be distri- butable between the two vowels, unlike that of the words first mentioned in this note, which gives nothing to the second vowel. NASAL REVERBERATION. /)1 apt to mistake for a syllabic sound), is made without any effort of those organs, being, in fact, but a sort of nasal reverberation.* But though vowels, mute in common or measured * A question may, perhaps, arise whether ' prison ' should be deemed a monosyllable or not. My own opinion is, that neither this word, nor any other, ending in ' on,' should be so deemed ; that the word is properly treated in Our prison strong : this huge convex of fire — , and shorn of its due proportions in Out of such pris'n, though spirit of purest birth — : and if it be urged that, ' risen ' being a monosyllable, ' prison ' ought to pass for one also, because both are often so sounded in common speech, I reply that in versification, as in other things, a line must be drawn somewhere, notwithstanding that the difference which exists in each side of the line ma}" be minute : the difference between tune, and out of tune, in musical instru- ments is often exceedingly minute ; yet it is not, for all that, a matter to be disregarded : no doubt, the ' o ' of ' prison ' is often dropt in common speech, just as Milton drops it in the verse above quoted; and so, also, is the ' o' (as Walker points out) of ' reason,' and ' treason,' and other words ending in ' on,' though I do not know of a case where Milton so treats it in any such word (save 'prison ') before an initial consonant ; while he often does so, on a quite different principle, before an initial vowel, as will be shown hereafter : but recited verse is declamation ; and declamation, as I have said, often brings out sounds apt otherwise to be lost : thus the syllables * trea ' and ' priz ' being taken, as they may be, separately, the final syllable ' zon,' complete in each case, remains, requiring distinct utterance ; and no one can deny that ' zon' (whether pronounced with the open ' o,' as in 'on,' or with the curt ' u,' as in 'under') has the elements of a complete syllabic sound : but in turning to words like ' heaven,' ' seven,' ' risen,' we find a different state of things ; for here the double consonant sound is not distributed ; it gives itself wholly to the prior syllable, and the final one gets no share : thus, though we say naturally ' sea-zon,' ' priz-zon,' we do not say naturally ' hev-ven,' ' sev-ven,' * riz-zen ;' so that if we make dissyllables of such words, there is no alternative between either adopting an artificial utterance, or ascribing syllabic honour to that which, as I have said above, is a mere nasal reverberation. And that the reader may understand what I mean, I would ask him to sound, first, the letter ' n,' as it is sounded in the alphabet (that is, with the vowel ' e '), and then endeavour to sound it without the vowel : the sound thus produced wiU be the nasal reverberation which I attribute to the words in question. In the words ' chasm,' ' prism,' 'rhythm,' 'phantasm,' this nasal effect is marked quite as decidedly as it is in the words before quoted ; yet becaiuse they are not spelt with a vowel intervening between the final and penulti- mate consonant, no one thinks of treating the three first as dissyllables, nor the last otherwise than it is treated in Me father, and that phantasm call'st my son. But if ' seven ' be a monosyllable, is ' seventh ' one also ? Milton makes no difficulty in so treating it ; as, for instance, in phrases like ' the sev'nth returns,' ' hallow'd the sev'nth day:' and this treatment must, 1 think, be pronounced admissible, notwithstanding that there results from it some harsh- ness of effect : on the other hand, I have nothing to say against using the word as a dissyllable ; for there is a clearly marked difference between the mere nasal sound of ' n ' and the sound resulting from union of that conso- nant with two other consonants. e2 52 A DOUBTFUL POINT. speech, should never be allowed to do service in verse, care must be taken to distinguish between those which are mute and those which are not ; for if we cut out, or slur,* a sylla- ble not bona fide mute, we commit an opposite fault, of which the effect will be more or less offensive according as the syllable so treated be naturally more or less distinct. With reference to this third rule, there remains yet to be considered another class of words ; namely, those having in them concurrent vowels of which it is not quite easy to say whether, or not, they ought to count in verse as separate syllables : such vowels are the ' ie ' of ' hierarchy,' ' society,' ' piety,' ' quiet ; ' the ' ei' of * deity,' ' being,' ' seeing ; ' the ' io ' of ' riot,' ' violence ; ' the ' ea ' of * real ; ' the double ' i ' sound of * flying ; ' the * ai ' sound of * saying ; ' and others of a cognate kind. Are ' hierarchy,' * society,' for instance, to be treated as words of four syllables, or as words of three ] are ' violence ' * diamond,' * diadem,' ' deity,' * piety,' to be held words of three syllables, or of two? are * being,' 'seeing,' 'quiet,' * riot,' to be held dissyllables, or monosyllables 1 It is a point towards the deciding of which Milton gives us no aid : the fact being, that, as usual, his authority bears, with about equal weight, on both sides of the question. | * To ' slur ' is to take no metrical account of a faintly sounding, but not mute, syllable ; it is often done by Milton, as is shown in the following lines of his, — A multitude like which the populous north — , Commended her fair innocence to the flood — , Wilt thou then serve this Philistine with this gift ? Of this practice, tlie best I can say is, that it is not so bad a fault as the converse practice : an undue curtailment of words in verse being, certainly, less offensive than an undue elongation. ' Aures enim, vel animus aurium nuntio, naturalem quamdam in se continet, vocum omnium mensionem . . . mutila sentit quaedam, et quasi decurtata ; quibus, tanquam debito frau- detur, offenditur ; productioria alia, et quasi immoderatius excurrentia ; quae magis etiam aspernantur aures ; quod, quum in plerisque, turn in hoc genere, nimium quod est offendit vehementius quam id quod videtur parum.' (Cic. Orat. liii. 178.) But according to Mr. Wm. Sydney Walker, our early dramatists used, metri gratia, to treat as dissyllables many words having three syllables each so distinct as the three in ' messenger,' ' Burgundy,' ' punishment,' and as monosyllables many others having two so distinct as those in ' flourish,' ' pro- mise,' 'barren,' 'forest,' 'Clarence,' etc. — (See ' Versification of Shakespear,' Art. IV. XXIV. XLIX. L. Refer also to p. 47, note *.) f This shiftiness of pronunciation is claimed by Dr. Newton for a merit COMPARISON OF EFFECTS. 53 Under these circumstances, our only course is to examine the two different modes, and decide for ourselves, as best we can, which of the two be preferable ; and if both shall seem admissible, then to determine on what principle such an ap- parent inconsistency can be justified. In the examples first about to be quoted, the two vowels count for but one syllable in the metre : — So sung the hierarchies ; meanwhile, the sun — , Variety without end : but of the tree — , From diamond quarries hewn, and rocks of gold — , By violence, no, for that would be withstood — , That gave thee being, still shades thee, and protects — , Epicurean, and the stoic severe— : < in these next, both the vowels are sounded and count : — His violence thou f earest, being such — , With diadem and sceptre high advanced — , That to the heighth of deity aspired — , My being gave me, or deceiv'd his heart — , The stoic last, in philosophic pride — , By owing, owes not, but still pays, at once — .* in Milton's versification. * Another liberty,' he says, ' that Milton takes likewise, for the greater improvement and variety of his versification, is, pronouncing the same word sometimes as two syllables, sometimes as only one, or two short ones. . . . but these excellencies of Milton's verse are attended with this inconvenience, that his numbers seem embarrassed to such readers as know not, or know not readily, when such elisions or abbre- viations of vowels take place.' Milton's two-fold treatment of these words may not be wrong ; yet if it be, as we are told, a liberty, there is nothing to be said for it : the excellen- cies of his verse are many'; but among them must not be ranked eflTects which cause embarrassment or inconvenience. * As 'voyage' (a word naturalised from the French) is invariably treated by Milton and Shakespear as a dissyllable, we might perhaps admit this to be a word of which the pronunciation has changed since Queen Elizabeth's time, were it not that Pope and Thomson, in the last century, both adopt the same pronunciation ; and there is much difllculty in believ- ing that any Englishman would have talked of a ' voi-yage ' so late as the year 1730. It may be, then, that these and other poets have given the word two syllables for no better reason than because it has two in French, or because they looked only to our manner of spelling. ' Seer ' is twice used by Milton as a dissyllable : by Shakespear, the word does not seem to have been used at all ; and I have no evidence to show whether or not Milton's usage accords with that of his cotemporaries. Some perhaps would treat it (though wrongly) as a dissyllable now : just as many so treat the word ' towards ;' the pronunciation of which I cannot, however, admit to be an open question ; though Milton evidently so 54 EITHER PERMISSIBLE. For my own part, I prefer (those cases excepted in which the close ' o ' predominates, as in * stoic,' ' owing,' ' growing') to treat the vowels as one syllable : at the same time, I cannot say that either mode seems to me necessarily better than the other, or either necessarily worse. For referring, once more, to what has been said touching the effect of declamation, we have to observe, firstly, that in all the words under review the tonic accent rests on the prior vowel ; and secondly, that though the succeeding vowel has a sound too faint for it to be counted as a separate syllable in ordinary speech, yet the two together do not make a sound in which either of them is lost ; but each, as in elision, re- *tains, more or less, its own distinctiveness. Let us take, for instance, the words * violence/ * diadem,' ' deity,' ' riot.' In treating the three first of these as dissyllables, and the last as a monosyllable, beyond doubt we do not pronounce them * vilence,' * didem,' * dety,' * rite,' that is, with the pure sound of our vowels ' i ' and * e ; * but in each case the faintly sounding vowel asserts itself as a qualification of the other ; is attracted towards it, but not effaced by it ; the * o ' of ' violence ' and ' riot,' and the * a ' of ' diadem ' impart to the tonic * i ' a certain open sound of the Italian ' a; ' and the * i ' of ' deity ' imparts to the foregoing tonic a sound which is not quite that of a prolonged ' e : ' now, if instead of sounding as one syllable the ' ia ' of * diadem,' * we prefer, in declama- considered it ; as is seen by the examples following ; the first pair giving the word used as a monosyllable, — Was moving tow'rds the shore ; his massy shield — , Of thunders heard remote : tow'rds him they bend — ; these next telling on the other side, — Straight towards heav'n my wand'ring steps I turn'd — , In serpent, inmate bad, and toward Eve — . I hold that in this word the ' w ' and the ' a ' are absolutely mute ; and that the pronunciation of it is either with the open sound of our ' o,' so as to make a parfect rhyme with ' lords,' or with the close ' o,' so as to make a perfect rhyme with 'fords:' treated as a dissyllable, it offends the ear almostasmuch as the word ' intere'sting,' which, instead of 'i'nt'resting,' we hear, every now and then, in the mouth of people who ought to know better. * This process of making one syllable out of two vowels whch do not, however, quite lose their own distinctiveness, is generally called ' synaeresis.' ' The Public School Latin Grammar ' distinguishes between the coalition of ' two into one long syllable or a diphthong ' (§ 12. pp. 15, 16) and the coalition of two into a ' quasi-diphthong,' forming one syllable. The first THE EEASON. 55 tion, to enforce separately the prior vowel ' i/ we are at liberty to do so, even though such a mode would have a pedantic effect in ordinary speech : for there is no limit (save that which strict time requii-es) to the enforcement of accent in declamation ; and a separate enforcement of the * i ' in * diadem ' does necessarily bring out, as a distinct syllable, the previously faint-sounding vowel * a ; ' thus * diadem ' becomes ' dii-ya-dem ; ' and the ear is not offended, because it perceives the change of effect to be wrought by a natural process operating on the bona fide elements of the words in question.* of these processes the author calls * synaeresis,' the latter 'synisesis:' but the ditterence between the two is hardly apparent : at any rate, it is not the effect of either process to make a diphthong ; which, as"^! understand the term, is a compound indivisible sound, resulting from the union of two dis- similar vowels. Vowels which suffer synaeresis may, not inaptly, be called a ' quasi-diphthong ;' but, being divisible, they cannot make a diphthong proper. In our language there are but three bonS, fide diphthongs, among the thirty-eight mentioned by Walker ; and these are, — 1. The ' ai ' sound, as heard in ' aye ; ' 2. The ' aui ' sound, as heard in ' voice,' ' boy ; ' 3. The ' ow ' sound, as heard in ' owl,' * pound,' ' crowd,' ' hour,' * power :' all the rest are either simple vowel sounds, or vowel and consonant com- bined. But when I speak of simple vowel sounds, it must be understood that besides the primary close and open sounds which pertain to all vowels, our ' a ' and ' u ' have other sounds : ' a,' for instance, has altogether four ; namely, the primary close sound, as heard in ' same ; ' the open sound, as heard in ' sad ; ' the Italian ' a,' as heard in ' art ; ' and the Saxon ' au,' as heard in ' all : ' then ' u,' besides its close sound, as heard in ' use,' and its open one as heard in * us,' has a further one, as heai-d in ' full,' ' should.' ' wolf,' ' bush,' ' Worcester ; ' for it is capable of proof that in each of these words there is identity of vowel-sound : and all these varieties of ' a ' and ' u ' must be deemed simple sounds, because it is not possible to show what are the component parts of any one among them that may be deemed compound. 1. The vowel sound of aye ' consists of the Italian 'a' and our close *e;' 2. That of ' voice,' * boy,' contains the Saxon ' au,' and our * i ' or * e ; 3. As regards the diphthong in ' owl,' ' pound,' etc., the best suggestion! can make is that it consists of our open ' o,' as heard in ' pond,' and our close ' u,' as heard in ♦ swooned.' Walker says it is composed of the * a ' in ' ball,' and the ' 00 ' in 'woo' (that is, the close 'u'), or, he adds, the *u' in ' bull :' but there is nothing of the ' a ' in these words, unless we pronounce them (as some do) ' aowl,' * paound ;' and, even then, it would not be the ' a ' in ' ball,' but the ' a ' in ' art.' Words like 'hour,' 'power,' etc., are best perhaps pronounced with a dash of the Italian 'a' in them. * Exactly the same thing happens in Italian, and other Romance lan- guages. In Italian, for instance, the two or more final vowels of words like 'mio,' 'tuo,' 'suo,' 'miei,' 'suoi,' 'via,' ' sia,' ' altrui,' and numberless 5Q OPEN VOWELS. HIATUS. I now come to the fourth rule : namely, that which for- bids a concourse, without elision, of final and initial vowels. Vowels thus meeting are called ' open,' because between them there is no intermediate sound caused by action of the vocal organs on each other, or by aspiration : the effect is an hiatus, or gap, which in all languages has been felt to be a blemish of verse ; and the effect of elision is to fill up the gap. Whether or not vowels were left open in Greek verse of the earliest times seems doubtful ; * but if Homer's hexameters were not free from hiatus, assuredly it found no place in the iambic verse of Greek tragedy. The Latins elided as a rule ; against which are of but others, always count for one syllable in the body of a verse ; but at the close of a verse tliey always count for two. Xom% if I be asked to give a reason for the uniform diversity of treatment which one and the same word receives, according to the position it occupies in verse, I have only to suggest that, the shorter mode seeming prelerable, Komauce poets have resolved, with one accord, to use no other in the body of a verse : but at the close of a verse (the metre being endecasyllabic) having to decide between not using words of this class at all, or treating their final vowels as separate syllables, they have preferred the latter course : and the alter- native has not embarrassed them, because the elements of distinct syllables are found, after all, to reside in the vowels. But if anj' one, going still further, require me to show cause why the vowels, while held to be separate syllables at the close of a verse, should be deemed inadmissible as such in any other position, I own myself unable to do so : all I can say is that, excepting now and then by Dante, they never are so admitted ; and that this uniformity of practice gives, in my judgment, a great charm to verse. * ' Hiatus is very common in Homer, and the other epic and elegiac, as well as lyric poets ; the tragedians admit it in the lyrical parts of their tragedies. . . . but never in iambic, trochaic or cretic verses.' Thus says Mat- ■ thiae (Gr. Gram., Blomfie d's Tr., p. 80) : but as re^ ards this alleged frequency of hiatus in Homer's verse, I would rather refer to a treatise on the subject which is to be found in the 7th vol. of the ' Classical Musaeum.' (No. iii., xxiv., xxxi.) The author (Mr, Bonnycastle) throws much light on points which others before him seem but little to have understood. His theory is this : agreeing with Bentley that the initial and consonant ^olic Digamma is practically the same as our consonant W, he asserts the same power to exist in other positions, whenever the vowels ' o ' or * v,' or their diphthongal compounds, are followed by any other vowel : but besides this power, or vis, of the W, there was also, he holds, present, and far more common, a sound of our con- sonant Y, whenever the letters ' e ' or ' i,' or their diphthongal compounds, were followed by a vowel ; and this sound he distinguishes by the term * Diijota.' It may be that Mr. Bonnycastle carries his theory too far : but certainly he makes out a strong case in favour of it up to a certain point. USAGE OF OTHER LANGUAGES. Ot small authority the few exceptions which may be quoted fi'om Virgil and others.* The Italians elide invariably where both the concurrent vowels are unaccented ; f and the cases are very rare in which they do not elide under other conditions. Such seems also to be the Spanish and Portuguese The French, following the precept of Boileau, Gardez qu'une voyelle, k courir trop hatee, Ne soit d'une voyelle en son chemin heurt^e, — profess to dislike hiatus in verse : but as they never elide any but mute vowels, we have yet to learn how they avoid hiatus between vowels which are not mute. J * In the whole iEneid there are, I believe, only twenty-one instances of vowels preserved from elision, and of these only three are preserved short, f The following lines from Dante, Lucrezia, Giulia, Marzia, e Cornelia — , Gente avara, invida, e superba — , Ma sapienza, e amore, e virtude — , contain instances of hiatus between unaccented vowels : but I cannot point to any other instances ; not even in Dante, whose verse, however, often pre- sents effects which are rarely used by others after him. I With reference to the line Hector en profita, seigneur, et quelque jour — , I find the following note in Geoffray's edition (1808) of Racine's works : "• Variante Hector en profita, seigneur, et en ce jour. "Get hiatus" (continues the editor) " ne se trouve que dans la premiere edition ; et il n'existe pas iin second dans les tragedies de Racine." ' The hiatus here specified occurs between two particles (' et ' and 'en ') ; and if M. GeoflFray means simply that, this one being removed, there remains not an example of the same kind throughout Racine's tragedies, the state- ment may pass unquestioned ; but not so,. if we are to understand it as pronouncing them free from all hiatus ; for examples occur in almost every page, e.g., Et si je viens chercher ou la vie ou la mort — , r.et enfant dont la vie alar me tant d'etats — , La reponse est dicte'e, et meine son silence — , Reste de tant de rois sous Troie ensevelis — , On m'envoie a Pyrrhus, j'entreprends ce voyage — , J'ai demande' These'e aux peuples de ces bords— , Ou s'est evanouie, ou s'est bi<^n relachee. It is vain to say that we are here dealing with diphthongs ; we are dealing with simple vowel sounds ; the alleged diphthongs being purely ocular : nor would the case be different if we suppose them real ; for a compound vowel- sound does not possess, any more than a simple one, the property of obviating hiatus. In writing thus, however, I do not at all mean to imply that elision is practicable in the cases before us. 58 ENGLISH USAGE. With us, the practice in this respect is lax beyond measure : yet the following extracts will show that eminent poets of our language have felt the evil, though they may have failed to contend against it successfully. Dryden says : ' Since I have named the " synalcepha," which is the cutting off * of one vowel before another, I will give an example of it from Chapman's Homer, which lies before me, for the benefit of those who do not understand the Latin prosodia ; it is in the argument of the Iliad : Apollo's priest to th' Argive host doth bring : — there, we see, he makes it not " the Argive," but " th' Ar- give ; " to shun the shock of the two vowels immediately following each other. But in his second " argument," in the same page, he gives a bad example of the quite contrary effect, — Alpha the prayer of Chryses sings, The army's plague, the strife of kings. In these words, " the army's," " the," ending with a vowel, and " army's," beginning with another vowel, without cutting off the first, which by it had been " th' army's," there remains a most horrible iU-sounding gap between the words. * I cannot say that I have everywhere observed the rule of this synaloepha in my translation ; but whenever I have not, it is a fault of soimd. * The French and Italians have made it an invariable pre- cept of their versification; therein following the severe example of the Latin poets. Our countrymen have not yet reformed their poetry so far ; but content themselves with following the licentious practice of the Greeks, who, though they sometimes use synaloepha, yet make no difficulty very often to sound one vowel upon another; . . . but it becomes us, for the sake of euphony rather " musas celebrare seve- riores," with the Romans, than to give into the looseness of the Grecians. 'f (Dedication of third Miscellany.) * Cutting off one vowel before another is not synaloepha (i.e. elision), but contraction. f It is probable that Dryden had not a very accurate knowledge of 59 * To come now,' writes Pope, * to the hiatus, or gap, be- tween two words which is caused by two vowels opening on each other (upon which you desire me to be particular), I think the rule in this case is either to use the caesura ' (he means elision) * or admit the hiatus just as the ear is least shocked by either; for (elision) sometimes offends the ear more than the hiatus itself, and our language is naturally overcharged with consonants ; as, for example, if, in the verse, The old have int'rest ever in their eye, we should say But th' old have int'rest ever in their eye. The hiatus which has the worst effect is when one word ends with the same vowel that begins the following; and next, those vowels whose sounds are neare-st to each other are most to be avoided. ... To conclude, I believe the hiatus should be avoided with more care in poetry than in oratory ; and I should certainly try to prevent it, unless the cutting it off is more prejudicial to the sound than the hiatus itself.' (Letter to Mr. Walsh.) Following the track of Pope, Cowper says, ' An alterna- tive proposes itself to a modern versifier from which there is no escape, which occurs perpetually, and which, choose as he may, presents him always with an evil : I mean in the instance of the particle " the." * When this particle precedes a vowel, shall he melt it into the substantive, or leave the Greek versification ; and an estimate of it, so disparaging as his, would hardly be accepted by the learned of our time. Nevertheless, a still more disparaging estimate was taught and accepted, till quite lately, in our schools, as the following extracts from the Eton Greek Grammar of 1856 show : ' Apostrophus est, cum eliduntur a, e, i, o et v, sequente dictione a vocah vel diphthongo incipiente : sed hoc, pro carminis ratione, vel observ- ant vel omittant Graeci ' (p. 183). Again, ' Hae sunt regulas observations dignissimae de syllabarum quantitate : sed quodammodo infinita est poetarum licentia, qui interdum longas corripiunt, vel breves producunt, metri necessi- tate coacti, aut ancipitem in eodem dictione, inque eodem versu, et pro- ducunt et corripiunt ' (p. 196). In the later edition, these passages are omitted. * Cowper' s perplexities are specifically confined to the definite article * the,' and so are, by implication, Dryden's and Pope's also ; but everything here said by him, or them, applies equally to all cases where final and initial vowels meet. 60 cowper's opinion. hiatus open ? Both practices ai-e offensive to a delicate ear. The particle absorbed occasions harshness; and the open vowel a vacuity equally inconvenient. Sometimes to leave it open, and sometimes to engraft it into its adjunct, seems most advisable : the course Mr. Pope has taken, whose authority recommends it to me : though of the two evils I have most frequently chosen the elision as the least.' (Pre- face to translation of Homer, pp. xxii-xxiii.) Now, if the alternative i-eally be as these authors state it, I cannot accept their conclusions ; for to me so offensive is hiatus, that nothing short of necessity can render it endura- ble : but if others, less sensitive, should deem, with Pope and Cowper, that elision, in some cases, is the worse evil of the two, then I would counsel that the difficulty be solved by choosing (since both are evils) not one of them, but neither. A poet is not bound (save when translating) to say any one particular thing ; at least, very seldom is he under that necessity ; nor again, is he bound to express him- self in any one set form of words : * if the words that first come to him be not suitable, let him seek others ; if others come slowly, let not that disturb him ; for no one heeds the throes of his travail, or is at all concerned to hasten the deli very. I * The difficulty immediately before us is confined to those cases in which *the' occurs before an accented initial vowel : to use it in such positions is, no doubt, unavoidable, at times ; and whatever is unavoidable must be sub- mitted to, whether we like it or not : but a poet who dislikes certain effects is at least bound to admit them no oftener than clear necessity requires ; and the number of cases in which there is a necessity to use the effect in question may be rendered very few even in a long poem. In the ' Seasons ' of Thomson (containing 5,420 lines), there is, I believe, but one instance of this effect, namely, in the line, Embow'ring endless, of the Indian fig : — and this one might easily have been avoided. Equally avoidable is the concourse of vowels found in the line, Of Nature, and the unimpassion'd shades : — where, probably, the same effect was intended, on an assumption that the first syllable of ' unimpassion'd' carries, or may cany, accent : but this is a ])alpable case of hiatus ; and is the only instance, known to me, in which Thomson forbears to elide ' the ' before an unaccented initial. t But of courise, if jieople have paid down cash in advance, they will ex- pect to receive the goods paid for within a reasonable time. On this princi- ple it happened to Dryden, after toiling three years over his ' Virgil,' that the clamorousness of subscribers forced him to bring out the work just four ELISION. CONTRACTION. 61 And here it is worthy of remark that the elisions and hiatuses, above cited by Dry den and Pope, are exactly similar, elision for elision, and hiatus for hiatus, to each other : thus Dry den adopts an elision which Pope rejects as inadmissible ; and Pope puts up with an hiatus which Dryden denounces as * a most horrible ill-sounding gap.' Still, they and Cowper are all of one mind on the main point ; all three acknowledging, in open vowels, a fault of sound, to be either excluded wholly from verse, or to be ad- mitted there but as a painful alternative. Now, the shock of vowels being found to be thus dis- agreeable, there is a constant tendency, in all languages, and in all speech, to take means for abating the nuisance ; and the means used are three, (1) insertion of an euphonic consonant, (2) contraction, and (3) elision. By us, the first of these means is used only in one or two cases ; the second very seldom ; the third almost every time we speak.* Contraction (which is the cutting off of a vowel or sylla- ble) occurs far oftener in the Romance languages than in English ; and on two points the difference of usage is specially to be noted. In those languages the definite article is always con- tracted before a vowel ; in English never. An Italian, for instance, says * V eccesso,' a Frenchman * I'exc^s ; ' but an Englishman does not say ' th' excess.' This rule is absolute. I hold, therefore, that Dryden and Pope both err when in the passages of theirs, above quoted, they write * th'Argives,' * th' old.' Again, the particles * di ' and * de ' are always con- Ssars sooner than he wished. (' Discourse on Epic Poetry,' p. 520. Ed. alone.) * As M. Jourdain spoke prose without knowing it, so those of my read- ers, who are not yet aware of the fact, may rest assured that, obedient to a hidden but irresistible law, they elide vowels incessantly. A mincing enunciation is the result of not eliding when two unaccented syllables meet : not only, then, is hiatus a fault of sound, but it offends, in such cases, against the natural tendencies of speech. Speaking of concurrent vowels in oratory, Quintilian says : ' Minima est in duabus brevibus offensio ' (L. ix. 4. 33) : the reason is, that weak syllables, thus meeting, naturally coalesce, so that, in fact, there is no hiatus between them. 62 ELISION DOES NOT ' CDT OFF.' tracted, before a vowel, in Italian and French ; while, in English, the corresponding particle * to ' never is : an Italian says ' d' inventar ; ' a Frenchman * d'inventer ; ' but an Eng- lishman never says * t' invent.' Different terms are used to express elision ; and there is hardly a grammarian among us who does not misstate its nature. But not to multiply examples, I will bere cite two authors ; Dr. Johnson and Dr. Carey.* The former defines synaloepha (i.e. elision) to be *a contraction, or excision, of a syllable in verse, by joining to- gether two vowels in the scanning, or by cutting off the ending vowel.' After quoting Maurus Terentianus, — Diphthongum aut vocalem haurit synaloepha priorem, Dr. Carey says : * Synaloepha cuts off the final vowel or diphthong of a word before the initial vowel or diphthong of the following word : f as, — Conticuere omnes, intentique ora tenebant — , Dardanidae e muris : spes addita suscitat iras ; in which cases we are to read, — Conticuer' omnes, intentiq' ora tenebant — , Dardanid' e muris, etc' Now, elision, as I understand the term, is a blending, or absorption, into one syllabic sound of two vowels concurrent in separate words ; and this is what seems to be meant by the ' haurit ' of Maui-us Terentianus : but a blending of two into one, and a cutting off of one from two, are things essentially different. I am not aware that there is any authority for applying this * cutting off' process to Latin verse | in recitation ; but * ' Latin Prosody,' f ' Elisions,' he sa^'s elsewhere, * are in general injurious to harmony, and the frequent recurrence of them is very disagreeable.' X Priscian, it is true, when he ' scans,' does cut off the prior vowels in such cases ; hut to scan a verse is to resolve it into its strict metrical ele- ments ; and in doing this, we rightly reject those parts which, as regards the metre, are inoperative ; in reciting verse, however, we have to proceed on a quite diflferent principle. EFFECT OF ELISION. 63 let US apply it to English verse. The following passages from Milton, — Less than archangel ruin'd, or the'^excess Of glory'^obscur'd — , A passage down to the'^earth, a passage wide — , contain elisions : now, suppose that, adopting Dr. Carey's method, we were here to cut off, instead of blending, and should recite the passages thus, — Less than archangel ruin'd, and th' excess Of glor' obscur'd — , A passage down to th' erth, a passage wide—, who would not stop his ears against such a jargon?* The effect of eliding the 'e' of Hhe' is, in fact, to con- vert it into the initial consonant * y : ' thus, * the excess,' ' the earth,' are sounded ' th'yexcess,' ' th'yerth ; ' and similar also is the effect of eliding the ' y ' of ' glory ' in * glory obscur'd. 'f Again, in the line. Pendant by subtle magic many'^a row — , ' many a row ' is really sounded ' men-ya-row,' * ya ' here making one syllable equivalent to the article ' a.' Thus two syllables are blended into one, while each re- tains a sound representing it in that one. Similar in principle is the effect of eliding the * o ' of * This method is actually adopted by Sheridan,with a view to demonstrate the bad effect of elision. Starting with the theory that EngUsh heroic verse often consists of more than ten syllables, he cites the line, And many a frozen, many a fier}' alp — ; and then, failing to distinj^uish between ' cutting off' and eliding, exclaims, ' what a monstrous hne this would appear if pronounced And man' a frozen, man' a fi'ry alp — , instead of that noble one which it is, when all the syllables are sounded.' Thus, one false theory leads to another : for though the line is monstrous if read with abscinded vowels, yet equally so is it if read by a process which not only yields three syllables too many, but grossly perverts the true order of pronunciation. t And yet, the customary way of printing ' the ' and ' to,' when held tq be elided, is, to cast out the respective vowels, and use the apostrophe. Thi^, at least, 1 find always done in the pages of Dry den. Pope, and Johnson. The right way to mark elision is to print the vowels with the apostrophe, thus, ' the' ' ' to" ' glory', ' etc. ; and this mode Dr. Newton always observes save in the case of ' the,' the vowel of which he cuts off', though why he should make this exception it is not easy to understand. 64 EFFECT OF ELISION. * to/ which changes, under the process, into the initial con- sonant ' w : ' thus, ' to invent,' ' to enter,' become ' t'winvent,' * t'wenter ; ' the difference between abscission and elision in such cases being as the difference between ' tin ' and ' twin,' between ' ten ' and the fiist syllable of ' twenty.' Similar likewise is the effect of eliding the final ' o ' of words like 'soitow,' * shadow,' as in Anguish, and doubt, and fear, and sorrow'^and pain — , * where ' sorrow,' followed by ' and,' becomes * sorr'wand.* In other cases of elision, it will be found that the concur- rent vowels form together a quasi-diphthong : by which term I mean any two vowels which, counting as one syllable, are yet not essentially indivisible. It is, however, to be noted, that owing to the peculiar manner of spelling which obtains in English, there is some- times an appearance of hiatus, without the reality : many words, for instance, spelt as though they begin with the vowel ' u,' do, in fact, begin with the consonant * y ; ' for words like ' use,' ' universe,' ' Europe,' are pronounced ' yuz ' (or ' yuce '), ' yuniverse,' ' yurope : ' again, the words ' one,* ' once,' begin, in fact, with the consonant ' w ; ' for there is no difference between the pronunciation of the numeral * one,' and the past tense of the verb * to win : ' before such words, then, no elision is needed ; and though, without it, there seems to the eye an hiatus, there is none to the ear. After acknowledging the music of the ancient tongues to exceed that of all now in use, and the poetry of Italy to be the most mellifluous of all modem poetiy. Dr. Johnson goes on to speak of elision as a license, f allowable, perhaps, in * Unaccented English vowel-endings consist, with very rare exceptions, of our nondescript vowel ' y : ' we have no such endings in ' a ; ' none (save of monosyllables) in ' e ; ' none, of words accented on the penultimate, in * i ; ' very few in it of words, like ' prophecy,' accented on the antepenultimate ; words ending in ' o,' like ' sorrow,' ' follow,' may almost be counted on the fingers ; and I cannot call to mind more than four words (viz. ' continue,' ' issue,' ' virtue,' ' value ') which, accented on the penultimate, end in ' u.' Our final ' y ' I caU nondescript, because it is really a thing quite apart, being unlike any other known vowel-sound ; and hardly capable of being distinctly sounded apart from a preceding consonant. f Dr. Newton also speaks of elision as a license (the term used is DR. JOHNSON ON ELISION. 65 some instances, but always, more or less, objectionable ; and one which, though used in many languages, ancient and modern, is unsuitable for ours. He then quotes from Milton the following lines, — No'^ungrateful food, and food alike those pure — , If true, here only,^and of delicious taste — , For we have also'^our ev'ning and our mom — , Inhospitably,'^and kills their infant males — , And vital virtue'^infus'd, and vital warmth — ; and adds, ' I believe every reader will agree that in all these passages, though not equally in all, the music is injured, and in some the meaning obscured. There are other lines in which the vowel is cut off, but it is so faintly pionounced in common speech that the loss of it in poetry is scarcely per- ceived, and therefore such a compliance with the measure may be allowed : Abominable,'^unutterable,'^and worse — , Impenetrablej^^impal'd with circling fire — , To none communicable'^on earth or heav'n — : * yet even these contractions increase the roughness of a language too rough already ; and though in a long poem they may sometimes be suffered, yet it can never be faulty to for- beai- them ' (' Rambler,' No. 88). Dr. Johnson, like others, seems to confuse between the cutting off and blending of vowels : for whatever may be the case as regards the three last examples quoted by him, assuredly in none of the first five are vowels cut off or lost. I deny, however, that elision injures music, or is more suitable to one language than to another ; on the contrary, regarding it as a means supplied by nature for neutralising a state of things which sensitive ears find painful, I hold that the reasons which are good for it ever anywhere, are good ' liberty ') ; but he tells us that Milton employs it as ' a method whereby to diversify and improve his numbers.' Other writers, like Sheridan, Thyr- witt, Sir Egerton Brydges, Mr. Steele, and Mr. Chapman, flatly assert that Milton does not use elision at all ! The two last-mentioned gentlemen, moreover, are of opinion that Virgil's lines may be made to sound better by ignoring it. 66 ELISION NOT LICENSE, BUT LAW. for it everywhere always : it is not, then, a license to be suffered, but a law to be observed ; the license is in forbear- ing it; which so far from not being faulty, is about the greatest fault a versifier can commit : * a strange plan it seems, for softening the verse of a rough language, to permit in it eiFects which others, not rough, abhor ; but stranger still is the reasoning that confutes the reasoner's own ad- missions; for if elision be licentious, inharmonious, and disagi'eeable, then not only must the Latin verse of Yirgil be owned very much to deserve these ill epithets, but the Italian verse of Tasso more than any in existence ; seeing that, on an average, the * ^neid ' has an elision in two lines out of every three, and the ' Gerusalemme ' near two in every line.f Elision occurs in the verse of Milton far oftener than in that of any other English poet : for while the rest scarce ever use it save with the definite article, and the particle ' to,' * That is, supposing the alternative to be between elision and hiatus, under the conditions heretofore specified. A poet, no doubt, is not bound to use elision unless he likes ; but not liking, he is at least bound to avoid a concourse of elidible vowels. f It is hardly, perhaps, necessary for me to say that I have not counted every elision in these poems ; nevertheless, the averages above given are founded on observations wide enough to ensure their accuracy: nor, in speaking of Tasso's versification, do I mean to say that elisions are more frequent in it than in Italian poetry generally ; the fact only is, that having tested the matter in his verse, I have not cared to test it elsewhere. We are told, indeed, by an editor of Cowper's ' Homer,' that when Tasso reconstructed his great epic poem under a new title, he adopted the rule of removing all elisions ; but how much truth there is in this statement may be judged by glancing at the two first stanzas : — lo canto I'armi'^e il cavalier sovrano, Che tolse'^il giogo'^alla cittk di Cristo : Molto col senno,'^e col invitto mano, Egli'^adoprb, nel glorioso'^acquisto : E di morte'^ingombrb le valli e^il piano, E correr fece^il mar di sangue misto : Molto nel duro'^assedio'^ancor sofferse. Per cui primo la terra'^e il ciel s' aperse. Quinci'^infiammar del tenebroso^inferno Gli'^angeli ribellanti,'^amori,^e sdegni, E spargendo nei suoi veneno'^interno, Contro gli'^armar del Ori'ente'^i regni, E quindi^il messagier del Padre'^Eterno Sgo'mbrb le fiamme'^e I'armi'^e gli'^odj^indegni : Tanto di grazia die, nel dubbio^assalto, Alia croce'^il figliol spiegata^in alto. — Gerusaleihme Conquistata. Milton's usage, elision and pauses. 67 and not often with tliem, he uses it in almost every case imaginable. Would I could add that he does so invariably, to the entire extinction of hiatus ! But consistency of practice, as I have said already, it is vain to seek in the versification of Milton ; and elisions and hiatuses are ever and anon found cropping up, promiscuously, side by side, in all parts of his poems. At this point it seems appropriate to consider what effect elision has on the phrasing of words in verse. If it be true that there is generally some interval between phrases not affected by elision, is there any between such as are so affected 1 if so, how do we reconcile it with elision 1 if not, how do we mark sentential stops between elided vowels 1 As the property of elision is to blend two into one, and of a pause to part two from each other, it follows that, theo- retically, there is no interval between the elidible vowels of separate phrases : practically, there is none between them in common speech ; and it is at least optional whether there be any in measured prose, or recited verse : the choice rests with the reciter; and so often as he chooses (having the power) to mark pointedly the grammatical close of vowel- ending words, there results, no doubt, some interval between the final vowel of the word so treated, and the initial one of the next : but such effects belong to elocution ; and the ear takes no offence at them,* at once perceiving that they are not caused by faulty arrangement. * The poet's part is to provide that the mechanism of his verse shall contain nothing which offends the ear ; and offence will inevitablv he given by any effect which, due to that cause, conflicts with the natural' tendencies of speech : now, concurrent weak vowels of separate words tend naturally to coalesce : the poet, therefore, is bound to see that thev are not left gaping in his verse ; but he is not responsible beyond : the reciter then steps in ; and if he, for purposes of his own, sees fit to keep the words apart, there is nothing to hinder him ; for he does but treat them according to strict grammar, which of course regards them separatelv : and the ear endures a severance in such cases, because instinctivelv it discriminates between the natural effects provided for by the poet, and those artfully introduced for a purpose which is beyond his province. The tendency of weak final and initial vowels to coalesce is marked most in those cases where the syllables immediatelv preceding and im- mediately following, f^re both accented, as in 'sorrow'^and pain,' 'the hol- low'^abyss,' where, indeed, it is irresistible ; still, it exists, and ought to be taken account of, under other conditions, as in F 2 68 ELISION AND PAUSES. Again, sentential divisions, represented by stops, are not, as I hold, necessarily marked by any distinctive suspension of vocal sound; nor do the rhythmical divisions of phrases necessarily coincide with the close of accented words : the final unaccented syllable of any such word may, without impropriety, run into a succeeding phrase; and whenever this happens, there takes place, between the accented syllable of the word and its unaccented one following, a pause which otherwise would take place at the close of the word. For instance, in the line, Sweet Auburn, loveliest village of the plain — : here, between 'Auburn' and 'loveliest,' there is a printed stop : now, some would phrase the line, — Sweet Auburn | loveliest | village | of the plain — , some, Sweet Au[burn, love|liest vill|age of the plain — ; and it cannot be said that either mode is wrong. However, if the former mode be taken, then there is a pause between ' Auburn ' and ' loveliest ' (though not a more marked one, I should say, than there is between ' loveliest ' and ' village ') ; if the latter, then there is no pause whatever between the final syllable of ' Auburn,' and the initial one of ' loveliest' (notwithstanding that a stop is printed, and a sentential pause deemed due) ; but there is a pause between the penul- timate and final syllables of ' Auburn ; ' and at that point, if anywhere, must be marked (if it need to be marked) the efifect indicated by the printed stop. It was admitted by implication, above, that in the last If true, here only,'^and of delicious taste — , InhospitablVj'^and kills their infant males—: in the lines following, To -whom its safety a whole nation owes — , His sword for glory, and his country's cause — , Undaunted truth, and dignity of mind — , the concurrent vowels are left open : and yet they would quite naturally unite if the words containing them were to occur, under similar conditions, in com- mon or measured speech. Even though the fact were otherwise, I should still contend that from such concurrences there results a fault of sound which ought to be avoided in vex'se. SYLLABLES WITHOUT VOWELS. 69 three examples given by Dr. Johnson there is a suppression rather than a blending of vowels. This occurs in the case of all words like * temple,' ' people/ ' battle/ ' humble/ 'circle/ 'abominable;' that is, of words ending with the vowel ' e,' preceded by the consonant * 1,' together with some other consonant : for instance. His temple right against the temple'^of God — , Arraying with reflected purple'^and gold — ; here, the final vowels of ' temple ' and * people ' are abso- lutely mute; and the consonants attach themselves to the initial vowels of the next word ; so that, to the ear, these are cases not of elision, but of contraction : in fact, the endings of such words are found not to be vowel-endings at all, if we consider what is the characteristic of a vowel-sound, namely, an efiect in the production of which there is no con- tact of the vocal organs : now, in pronouncing the final syllable of these and similar words, it will be found invariably that the tongue rests on the palate when the efibrt of pro- nunciation ceases.* Contracted, likewise, are final syllables in ' er ' and ' on ' "f * On this point Walker well remarks : ' L, preceded by a mute, and fol- lowed by an " e," in a final syllable, has an imperfect sound which does not much honour our language. The L in this situation is neither sounded "el" nor "le," but the final "e" is suppressed, and the preceding mute articulates the L, without either a preceding or succeeding vowel : so that the sound may be called a monster, in grammar^ — a syllable without a vowel. This will easily be perceived in the words '* able," " table," " circle," etc., which are pronounced " abl'," "tab!'," "circl'," 'etc. Even so far back as the year 1653, we find Dr. Wallis treating as mute the final ' e ' of such words, e.g. : ' Quando autem nulla praedictarum rationum urget continuationem ipsius e, ab accuratioribus typographis nunc dierum omittitur : nisi quod post I, alii consonae subjunctum, a plerisque adhuc retineatur, ut in " candle," " handle," "little," " wrangle," "possible," etc., in quibus nulli nunc inservit usui, adeoque non incommode omitti poterit (' Gramraatica Linguae Anglicanae,' Ed. 1). f Contractions of the mute final ' e ' and of the final ' er ' and * on ' occur often in the verse of our early dramatists ; as is shown by Mr. Wm. Sydney Walker, in his ' Versification of Shakespear ; ' (see from p. 67 to p. 74) and among the many peculiarities he sanctions, these are the only ones which to me seem admissible. As regards the final * er ' and ' on,' such effects take place by operation of the same law as that which, in the body of a word, renders mute an unaccented vowel, when followed by the consonant ' r ' or 'n:' thus, 'sever, sev'rance ; ' 'savour, sav'ry ; ' 'ever, ev'ry ; ' 'soften, soft'ning;' 'reason, reas'ning ; ' and, in like manner, we have 'savour, sav'rof ; ' ' river, riv'rof ; ' ' soften, soft'nand ; ' ' reason, reas'neth.' 70 FINAL ' ER ' AND ' ON.' (however these sounds be represented to the eye) when they occur under conditions observable in the examples foUow- And where the river'^of bliss through midst of heav'n — , Before them in a cloud or pillar'^of fire — , All judgment whether'^in heav'n, or earth, or hell — , The savour'^of death, from all things else that live — , Whom reason'^hath equall'd, force hath made supreme. Under different rhythmical conditions, this kind of con- traction, though still possible, suggests itself less obviously : for instance, if we suppose lines such as A river in the midst between them flow'd — , And a pillar of fire before them went — , the words here adapt themselves to the rhythm without con- traction ; and have not, in the absence of it, a bad effect. But the imperfect endings of words like ' temple,' ' people,' have not the same latitude of adaptation : followed by an initial consonant, they necessarily, indeed, pass, and followed by an accented initial vowel, they may perhaps pass, for independent syllables; but they lose that character under different conditions : for the natural tendency of consonant articulation is to unite with vowels which do not repel union ; and whenever, therefore, a weak syllable, beginning with a vowel, succeeds any such imperfect final sound, effect is given, or is due, to this tendency.* The right mode, in my opinion, to print words ending in * er ' and * on,' Avhen they suffer this kind of contraction, is to cut out the vowel, and use the apostrophe, thus ; ' riv'r of bliss,' ' pill'r of fire,' etc. Dr. Newton, after saying (note 1, p. 248) that in such cases the word is to be pronounced ' as one syllable or two short ones,' uses a manner of printing which is con- sistent with neither treatment, namely, ' riv'er,' * piU'ar,' ' reas'on,' etc. If the word is to be treated as one syllable, the vowel ought to be cut out, and the apostrophe used ; if as two syllables, there is no need of the apostrophe. * Here, also, as in cases of elision, the tendency to coalesce, though marked more strongly when both the adjoining syllables are accented, ex- ists, however, when these conditions are wanting : thus, in the line, Swift as the sparkle of a setting star — , the imperfect syllable is made to count in the metre ; but if, metre apart, the phrase ' sparkle of a setting star ' were to occur in common or measured speech, the final syllable of 'sparkle' and the particle 'of would make together but one sound. 'he,' 'his,' 'have,' not always aspirated. 71 The lines To'^whom Satan, turning boldly, thus replied — , But he, the Supreme Good, to^whom all things pure — , also present cases of contraction : for although the * w ' of * whom ' is always mute, and the ' h ' often so, in common speech, yet it never is after the particle ' to : ' thus, we may say, without aspiration, * for oom ; ' but we cannot say, with- out it, ' to oom ; ' and the * h ' aspirate is a bar to elision. This, therefore, must be considered a case of arbitrary con- traction, like ' I'll ' for * I will,' sufficiently accredited to be admissible in vei-se ; and I call it arbitrary, to distinguish it from the contractions heretofore just mentioned, which take place by a natural process. And as after a consonant we often drop the initial aspira- tion in ' whom,' and ' whose,* we drop it also after a con- sonant or a vowel, in the personal and possessive pronouns * he ' and ' his,' when they precede, respectively, their verb or substantive; so that elision takes place naturally between them and any preceding unaccented vowel. In fact, these pronouns, together with the auxiliary verb ' have,' * are found to take, or to reject, aspiration, and therefoie to be elidible, or not, exactly according to the conditions already stated (pp. 10, 11, 12), which cause them to be accented, or unaccented : e.g.. Concerning thee to'^his angels : in their hands — , This city'^his temple, and his holy place — , Not this place only'^his omnipresence fills — , Deposited within me, which to'^have kept — , Worthy to'^have not remain 'd so long unsung — , Nor should'st thou'^have trusted to that woman's frailty — : hence, the line, This universe we have possess'd and rul'd — , is faulty ; f because either there is hiatus between * we ' and * Even the possessive verb 'have ' is not aspirated, when pJaced in sub- ordination, as, for instance, in Love seeks to have love ; where elision takes place between ' to ' and ' have.' f A similar ill effect there is in the line Of him who had stole Jove's authentic fire — 72 DIGAMMA. * have,' or there is attributed to * have ' an aspiration not here due. On the other hand, the lines I had continued happy, had not my fault — , Of high collat'ral glory. Him thrones and pow'rs — , are faulty in the opposite direction; because between * happy' and 'had,' between * glory' and 'him,' elision is attributed ; notwithstanding that both ' had ' and ' him ' require aspiration. As our language has a large number of dissyllables and monosyllables ending with an accented vowel, or diphthong sound (sueh as ' say,' ' away,' * decree,' ' sea,' ' sigh,' ' eye,' * descry,' * know,' * bestow,' * to and fro,' ' few,' ' renew,' 'joy,' ' alloy '), it is of importance to decide the rhythmical quality of such sounds ; that is, whether or not, they present, when followed by a vowel, the alternative of an hiatus, or a harsh elision : to me it seems that they do not ; because from such concurrences in our language there is always found to result an intermediate sound of either the consonant ' w ' or of the consonant ' y.' In fact, these final vowels are digam- mated.* And here again, if there be any question, let an experiment be made : let the objector try if he can sound, naturally, in measured recitation, any one of these words, before an initial vowel, without some contact of the vocal organs : in the case of words ending with an ' o ' or ' ou ' sound, he may have ocular proof by making the experiment before a glass. The verses following furnish examples in which the opera- tion of this law may be observed ; — a line which has much perplexed the commentators ; though no one has ventured to pronounce it (what really it is) faulty : for here, the metre requires a separate syllabic utterance of ' who ' and ' had,' while the proper rhythm of language' requires elision, which in the following line from' Sam- son Agonistes ' As vile had been thy folly, who'^have profan'd, takes place. In such cases, then, the ear expects, and takes account of, eli- sion, and if this be not provided for, the result is that the Une seems to halt, lacking a syllable. * The term 'Digamma' is applied in this treatise alike to the latent power of our ' w,' and to the Diljota, or latent power of our ' y.' The term ' Diijota' is borrowed from Mr. Bonnj'castle. (See p. 56, note *.) 'the' and 'to.' 73 Scowls o'er the darken'd landscape snow, or shower — , The birds their notes renew, and bleating herds — , Fly to and fro, and on the smoothed plank — , As when, far ofE at sea, a fleet descried — , But who are these ? For with joint step I hear — . But if the accentuation of a final vowel, in presence of an unaccented initial one, be preventive of hiatus in our language, the question arises whether the accentuation of an initial vowel, in presence of an unaccented final one, does not have the same effect. In the two first books of * Paradise Lost,' there are about forty-nine instances of vowels left unelided under the last- mentioned conditions; and if these be classified and examined, there will be found occurring before accented initial vowels, The particle < the ' 22 times, The particle « to ' 9, The pronouns ' he ' and * she ' 2, The preposition ' by ' 4, The pronoun ' thou ' 1, The conjunction ' through ' . 1,- The final ' y ' as in ' easy ' ... 10. 49. Thus, in thirty-one cases out of the forty-nine, that is, in nearly three-fourths of the whole number, we have to deal with the particles * the ' and * to,' of which neither can be inflected or cui'tailed. But here a question arises. Since our language, unlike most others in analogous cases, never cuts off the vowel of these particles, must there not be a reason for a state of things so exceptional? if Pope would not bring himself even to elide the ' e ' of ' the ' before ' old,' may not such re- pugnance on his part be traced to some cause other than the supposed ill effect of blending together an unaccented and an accented vowel 1 Again, the vowel- sound of our particle * the ' is here just the same as that of the Italian ' gli,' which (except before words beginning with ' i ') is never contracted, though 74 VARIABLE VOWEL-SOUND OF 'THE.' always elided : the elision, then, of * the ' before * order,' would correspond with that which takes place in the follow- ing line of Tasso, De soave licor gli'^orli del vaso : now, elisions of this kind occur, as a matter of course, in almost every page of Italian poetry, which is admitted, nevei-theless, to be the most melodious in existence. Why, then, should a state of things which causes no unpleasant- ness in one language, be held to cause it, under conditions which seem exactly similar, in another? The answer, I think, is, that the conditions are not quite so similar as they seem. The sound of our article ' the,' before a consonant, is one which cannot exactly be represented by any vowel or diph- thong; but it is identical with the vowel-sound of the French particle ' le : ' now, if this sound were retained be- fore an accented initial vowel, it would indeed cause a cacophony which might well be called (to use Dry den's words) * a most horrible ill-sounding gap : ' our ear, there- fore, at once rejecting it, ascribes to the article, in such posi- tions, the true vowel-sound of our letter ' e,' as heard in the pronoun ' thee ; ' and thus we say ' thee earth,' ' thee air,' ' thee oracle,' ' thee overture,' ' thee anchor,' etc. Two articles (' a ' and ' the ') are evermore, and irrepres- sibly, recurring in our speech : accordingly, we have to say ' a king,' ' a road,' ' a mountain : ' but we cannot say ' a ant,' * a anchor,' ' a obstacle.* Why not 1 Because of the hiatus :* so to get rid of the hiatus, we convert ' a ' into ' an,' and thus ends all difficulty with respect to this article. In like manner, when dealing with the definite article, we say ' the road,' ' the king,' ' the mountain ; ' but as we cannot, because of the hiatus, give this article, before a vowel, the same sound it has before a consonant, here also we have recourse to an expedient; but the expedient here consists, not in * Strange though it may seem, the fact, however, is, that our three par- ticles ' the,' ' a,' and ' to,' all take, before a consonant, the above-mentioned vowel-sound of the French particle ' ie.' INTERMEDIATE Y-CONSONANT SOUND. iO taking up an euphonic consonant, or in cutting off an obnoxious vowel, but in changing, as I have shown, the vowel sound : which done, we stop. But why do we not, having the power, go on? why do we rest satisfied with the change made, when a further change is quite practicable 1 why not boldly cut out the ' e ' of ' the,' as the French and Italians do the corresponding vowel of their definite article 1 Because our ear tells us that from the newly imparted vowel- sound a consonant sound results; and that enough, therefore, has been done to fill up the vacuity. Now, if anyone will repeat in succession, so as to make a sort of scale, all the various initial vowel-sounds, such as we have in ' author,' ' army,' ' angel,' ' anchor,' ' eagle,' 'entrance,' * idol,' ' image,' ' omen,' * obstacle,' * owl,' ' Ouse,' ' utterance,' prefixing to each this particle ' the ; ' thus. the au, the im, the a (Ital.), the 0, the a (Eng.), the ob, the an, the ow (as in owl), the e, the ou (as in Ouse), the en. the utt. the i. he will find (xmless pains be taken to prevent it) that an intermediate sound of the consonant * y ' makes itself per- ceived throughout;* and what I suggest is, that to an innate, though, perhaps, unconscious, perception of this sound's pre- sence, is due alike the non-contraction of our definite article. * In the rapid utterance of common speech, this sound may be lost ; but in the more slow, more strongly marked, more distinct utterance of declama- tion, it is either heard, or may be heard. I do not deny that even in declamation, the article may be uttered before any of the above-mentioned words, without producing the ' diijota : ' what I assert, however, is, that when not elicited, the sound is still latent, and may be elicited at pleasure, because it is an effect resvilting naturally from the vowels here concurrent, when rendered in a way which does not go beyond the privileges of declaimed speech. The utterance which marks the ' diijota ' differs from that which does not mark it, only in a more com- pletely distinct enunciation of the pure sound pertaining to the vowel ' e.' Both modes are natural ; and it is impossible to say that one is more natural than the other : but one is more suitable for recited verse ; and a choice of either being offered, we are at liberty to take the one which best suits our purpose. 76 Y-CONSONANT SOUND. and the repugnance, felt by Pope and others, to elide it, in certain cases. And if I be told, by way of objection, that, according to this theory, there will practically be no difference between ' arrow ' and ' Yarrow,' • between * ear ' and ' year,' ' east ' and ' yeast,' ' earning ' and ' yearning,' * oak ' and ^ yoke,' and so forth, I reply, that though between such words, under other conditions, there is a well-defined diversity of sound, yet, between them, when preceded by the definite article, the difference is hardly, if at all, perceptible. And here the reader will note that I have not asserted the* diijota' to be entirely identical with the consonant ' y : ' I would rather describe it as a * vis,' a power, of that letter, rather than the full power : and if I be asked to illustrate the distinction thus drawn, I would instance the effect this letter has in * yoke,' ' year,' ' yeast,' and * yearning,' when these words, or the like of them, are uttered separately, or after a final con- sonant, as compared with the effect it has in them after the definite article : in the latter case the * y '-consonant sound, though still heard, is heard somewhat less forcibly than be- fore; and in this less forcible utterance we have just the state of things which brings us to the * diijota.' Now, words like 'oak,' * ear,' ' east,' 'earning,' etc., take, under similar circumstances, a sound, though not, it may be, the full sound, of the letter 'y:' on the one hand, then, a previously existing sound is reduced to a certain point ; on the other, a sound not previously existing, asserts itself up to a certain point ; and thus, by a simultaneous process of ap- proximation, we arrive at results which, to all intents and purposes, are identical in effect. Indeed, so close is the ap- proximation, that the difference remaining, if any there be, is very much the same as that which exists in music between G sharp and A flat; notes which, though theoretically separated by an enharmonic interval, are, nevertheless, represented by one and the same sign on keyed instruments. And to test the truth of what I here advance, let the reader try the experiment of eliding ' the ' before an initial 77 * y,' side by side with elision of it before an accented initial vowel ; and he will find, if I mistake not, results all but identical : and if I be told that there can be no elision before a consonant, my answer is that the initial ' y,' while possessing, in certain positions, the full power of a consonant, is yet found to have, in other positions, some property of a vowel : for not only, as is shown above, does it take, when preceded by the definite article, a sound less distinct than the one it takes when standing singly, or coming after a conso- nant, but * the ' itself, when followed by this letter, takes, or may take, very much the same sound it has when followed by an accented vowel.* Such are the reasons which occur to me as sufficient to justify us in exempting from the reproach of hiatus all these rhythmical phrases in which * the ' is used, unelided, before an accented initial vowel, f * ' Y,' when it follows a consonant, is a vowel ; when it precedes either a vowel or a diphthong, it is a consonant : ' ye, young.' It is thought by some to be in all cases a vowel : but it may be observed of ' y ' as of ' w,' that it follows a vowel without any hiatus, as in ' rosj' youth.' (Dr. Johnson's ' Grammar of the English Tongue. ') Undoubtedly, as Dr. Johnson says, there is no hiatus in * rosy youth ; ' nor, as I should say, in ' the youth ; ' but equally, to my mind, beyond doubt is it that the initial * y ' of ' youth ' is not heard, in either case, as it is heard when the noun stands alone, or is preceded by a consonant. The following line from Ramsay's ' Gentle Shepherd,' My father's hearty table you soon shall see, is remarkable, as containing a contraction (and, I think, a perfectly legiti- mate one) of the final syllable of ' table ' into one syllabic sound with ' you :' now, such an effect would not occur before any consonant other than a * y.' Even in Milton's line Justly, yet despair not of thy final pardon (' S. A.'), elision between the final vowel of 'justly,' and 'yet,' seems an eff'ect quite admissible : and that Milton so intended it I believe for this reason : if eli- sion is not admitted, there is a syllable too many in the verse ; and I find not so much as the appearance of such a thing elsewhere in * Samson Agonistes.' The French word ' yeux ' affords a further remarkable testimony as to the double character of this initial ' y : ' the ' y ' of ' yeux ' is just as much a consonant as the ' y ' of ' youth,' or of any other English word beginning in the same way ; and yet, nevertheless, in ' les yeux ' the ' s ' of ' les ' is sounded as though it were followed by a vowel. t The vowel-sound of the Italian ' i ' in * gli,' being identical with that of ' e ' in ' the,' and the ' o ' of ' orli ' with that of ' order ' (see p. 74), it may be urged that if the digamma be present in ' the order,' it will be present, likewise, in ' gli orli.' All I venture to say on this point is, that, theoreti- cally, there seems no reason why this power should not exist, under the conditions supposed, in Italian, provided Italians choose to pronounce their 78 WHEN, AND WHY, ' TO ' NEED NOT BE ELIDED. The particle ' to,' which has next to be considered, has. also a variable pronunciation : on the one hand, it takes, like * the,' before a consonant, the vowel-sound of the French particle ' le ; ' on the other, it takes, before a vowel, not the true sound of our vowel * o,' but that of our vowel * u,' as heard in ' you ' and ' through,' which beyond doubt do not lend themselves to elision before accented vowels : now, these words are sometimes accented, sometimes not ; but ' through ' is certainly not accented when, without being preceded by a verb, it is used as a preposition immediately before a noun- substantive, as in the line, Through all the changing scenes of life ; a line which few people, I think, will be inclined to pronounce faulty : but if there be no hiatus between ' through ' and ' all,' I do not see on what principle there should be held to be any between ' to ' and ' all ' in the line To all you ladies now on land : I hold indeed that there is none in either case ; and that the argument above applied as to the presence of one form of digamma, applies equally here as to the presence of another : nor is ocular proof in this case wanting, for as sm-ely as the hammer of a pianoforte strikes the string above it whenever the corresponding note on the key-board is touched, so surely, m these and like cases, do the lips, by closing, testify that a consonant sound is uttered. Granting then that the final unaccented ' e ' sound of ' the,' and the corresponding ' u ' sound of ' to ' are respectively digammated in the positions aforesaid, it will follow that many similar endings are, in such positions, aiFected after the same manner ; and similar, on the one side, is the ' e ' sound of * he,' * she,' and ' ye ; ' and similar, on the other, are, as I have shown, the ' u ' sound of ' you ' and ' through,' which. definite article, before an accented vowel, with as much distinctness as we pronounce ours ; but, in fact, they do not: for their invariable rule is to con- tract ' gli ' before all words beginning with an ' i,' and to elide it in all other SAME LAW APPLIED IN OTHER CASES. 79 indeed, are the only unaccented monosyllables in our lan- guage having a vowel sound exactly like that of our ' to.' But there are other vowel endings which, on the same principle, are similarly affected also ; and these are our ' i ' sound, as heard in ' by,' and the pronoun * I,' and the primary close sound of our 'o,' as heard in 'so,' and ' though ' (which are often unaccented), and the * ow ' diphthong sound, as heard in 'thou.' Thus, the line, Knowing who I am, as I know who thou art, which, at first sight, seems to have three hiatuses, is found on examination to have none. As regards the final ' y ' of words accented on the penul- timate or antepenultimate, I have only to say that it does not, in my judgment, carry 'with it any digammating power.* * This letter has two distinct sounds ; one, as heard in * prophecy,' ' tes- tify,' which is the pure primary sound of our vowel 'i ; ' another, as heard in ' plenty,' ' secrecy,' ' fallacy,' of which I scarce know what account to give, as it cannot be referred to any other known vowel-sound, being, by itself, a sort of open grunt, at once guttural and nasal. Ben Jonson's statement that compounds of 'facio,' like 'liquefy,' ' qual- ify,' &c., are accented on their final syllable, is not (as I believe) founded on any rule of pronunciation peculiar to his time, but on a false assumption, still common among grammarians, that accent, in certain cases, is a neces- sary accompaniment of vowel-sound. A remarkable instance of this error is furnished by Dean Alford. After quoting a correspondent who complains of the stress laid (as he says) on the final syllable of ' prophecy,' and who asks what we should think of ' extasy,' ' fallacy,' or ' phantasy,' especially if put in the plural, the Dean goes on to say : 'In this case usage is right, and apparent analogy wrong ; " extasy," as we have already seen, is from the Greek word "exstasis," "phantasy" from the Greek word "fantasia," "fallacy" from the Latin word "fallacia;" but "prophecy" is from the Greek " profeteia," and it is, therefore, not without reason that we lay the stress on the last syllable. The verb to " prophesy" we pronounce in the same way ; I suppose by a double analogy : partly guided by the sound of the substantive, partly by that of other words ending in "y," as "qualify," " multiply," "mystify."' (' Queen's English,' pp. 63, 54.) Now, observation should have taught Dean Alford that stress, or absence of it, on syllables of English words derived from Greek or Latin, is very little indeed dependent on the question of ' long ' or ' short ' in corresponding syllables of those languages : there is, then, no relevancy in attributing to Greek or Latin origin the syllabic sound of this or that English word, unless it can be shown that all syllabic sound in English words of like origin is to be ac- counted for on the same principle. So long as we are content to take our lan- guage as it is, to study its effects, and make the best of them, we stand on safe ground ; but if people must needs render a reason for everything they meet with in the course of such researches, they will soon find themselves involved in all sorts of inconsistencies and contradictions. But Dean Alford and 80 *CiETERIS PARIBUS.' Now, although the arguments here used, and the con- clusions come to, may seem at variance with the remarks made in p. 60, yet there is in reality no contradiction : for there I argue, on the assumption of eminent poets, that hiatus results from certain combinations of words, whereas here my argument is, that from those, and similar combina- tions, no hiatus results : the sole question with me being, not whether there be less evil in leaving vowels open than in eliding them, but whether, elision not being used, the vowels are left open or not : positively, I maintain that if hiatus be a fault of sound, it is just as much so, cseteris paribus, in one language as in another ; * and not the least do I incline to Dr. Johnson's dictum, that elision, while suitable for other languages, may be unsuitable for ours, unless we can allege a difference of conditions, and point out wherein the difference consists. But though Milton very often does not elide before ac- cented vowels, he very often, however, does ; as is seen by the following examples, — Before all temples the'^upright heart and pure — , Of tow'ring eagles to'^all the fowls he seems — , The'^earth cumber'd, and the wing'd air dark with plumes — , A passage down to the'^earth, a passage wide — , Now morn her rosy steps in the'^eastern clime — , his correspondent are both on the wrong track : they fail to distinguish be- tween two things essentially distinct ; between stress (which is accent) and mere vowel-sound. The only difference between the final ' y ' of ' ecstasy ' and that of ' prophecy,' is, that, in the former case, the vowel takes the non- descript ?ound above mentioned, and in the other, the pure vowel-sound of our ' i ; ' but one vowel-sound is not, by its nature, a jot more liable to stress than another ; nor, in fact, is the ' y ' of any among the words before us, entitled to the least stress. * I will here point out an inconsistency which has often struck me. Let any Englishman, who may be fitted for the task, undertake to examine some fifth-form Latin verse exercise, and he will deem intolerable the slightest flaw in the versification. How much does an hiatus pain him ! what sen- sitiveness of nerve he shows if a single false quantity be committed ! But let this rigid uncompromising critic turn then to Englisli verse, and we shall find him unhesitatingly accepting, or, it may be, himself using, the very- same effects, or the same in principle, as those by which he had just before been so much shocked in a school-boy's crude effusions. Xow a Latin hiatus is no worse than an English one ; and the principle which forbids the use of false quantity in an ancient tongue, applies with equal force against the use of false accent in a modern. OF OPPOSITE MODES NEITHER MAY BE WRONG. 81 Laden with fairest fruits that seem'd the'^eye — , Chary bdis, and by the^other whirlpool steer 'd — . As from the centre thence to the'^utmost pole — , Into'^utter darkness, deep ingulf 'd, his place. Now, it seems, at first sight, impossible that each of these opposite modes can be right : nevertheless, in this case, as in others, we shall find it not easy to pronounce either wrong. The key of the difficulty will, I think, be found, if, while recognising the power or ' vis ' of the ' y ' and of the ' w,' we take account of their twofold operation. I have already shown that when ' the * or * to ' are elided, the effect is to convert the vowels ' e ' and ' o,' respectively, into the initial consonants ' y ' and ' w : ' the immediate cause of this conversion is the rapidity with which the particles are pronounced : thus, whenever the vowels of these particles meet other vowels, the consonant ' vis ' is always present and operative, though not always operative as a digammating ' vis :' when the initial syllable is unaccented, elision natu- rally takes place, and we see what then becomes of the final vowels ; but if there be accent, it causes a resistance which retards, or may retard, the voice, and, retarding it, brings out the digamma : but still, elision is practicable, provided the particles be uttered with a rapidity sufficient to overcome the resistance offered ; and we can so utter them, or not, at pleasure, because there is no law which prescribes how fast, or how much less fast, the utterance should be : but the choice * is open only where there is accent ; for it is resistance which brings out the digamma; and unaccented syllables offer no resistance, f * Diversity of practice, under seeming identity of conditions, on any given point, is not, I must own, a thing to be wished for in verse ; and I would gladly devise some principle for the guidance of versifiers in exercising the discretion which seems due to them on this point : but none such occurs to me ; nor do I believe that a reliable one is to be found. t The only point on which I incline to differ with Mr. Bonnycastle (see p. 56, note *) is in doubting whether these consonant powers of the 'w' and ' y ' would make themselves felt in Greek when each of the consonant vowels is short, as in a-Vve' lfli?/ce : if it would, then we must suppose the ancient Greek to have been more strongly digammated than our English is : and perhaps it was To me, hoAvever, there seems a greater probability that in this and similar cases the gap was filled up by some other process. G 82 TASSO ON CONCURRENT VOWELS. One reason, as I have said, why open vowels are objec- tionable in verse, is, because there is a natural tendency to blend into one syllabic sound the final and initial vowels of separate words : but sometimes (as when, for instance, both vowels are accented) this tendency does not exist ; so that, in such cases, the objection above stated does not apply ; and the question arises whether concurrent vowels, which do not easily coalesce, are admissible in vei-se without elision 1 In the prose writings of Torquato Tasso,* there is a rather remarkable passage, bearing partly on this point. Speaking of several things which cause ruggedness (asprezza) of composition, but, at the same time, a certain grandeur and gravity, he quotes, in illustration, four verses of Dante, ' in which,' he says, * the vowels are not absorbed, but there is made, as it were, a gap and a chasm : ' — Pol ^ Cleopatras lussuriosa — , La onde il carro gik era sparito — , Vid' lo scritte al somma d' una porta — , Nel del che piii della sua luce prende, F'u i'o — etc. : * although,' he adds, 'the concourse of i does not cause so great a chasm or hiatus as that of a and o, for which we are wont to open the mouth wider.' Now, such concurrences being prima facie licentious, while Tasso here speaks of them as effects permissible, nay praise- worthy, at times, it is well worth while to pass in review each of the examples thus approved by him. Pol h Cleopatras lussuriosa — : 'poi,' which in the body of a verse should count for but one syllable, is here used as two ; and no elision takes place between the final ' i ' and the ' h ' which follows. Vid'io scritte al somma d' una porta — : here, too, the ' io,' which ought to count for but one syllable, counts for two ; but this is not a case of hiatus in the ordi- nary acceptation of the term. But as nothing which Tasso Del Poema Eroico,' Libro Quinto. 83 writes on matters of versification is to be passed over lightly, I will here pause to enquii-e why it is that neither of these two effects have seemed offensive to his ear. He admits that the concourse of the ' i ' does not cause, in these cases, so great a gap as that of the * a ' and * o ' in ' la onde ' (and he might have added as that of the 'a' and *e' in *gia era') ; and the reason, he says, is, that the mouth is not opened so wide in the one case as in the other : but is there not, de- pendent on this, another reason, which he fails to mention, namely, that in passing from the * i ' of * poi ' to the accented * d,' and from the tonic * i ' of * io ' to the unaccented ' o,' — in both these cases, the tongue touches the palate, and so pro- duces the y-consonant sound 1 So, at least, to me it seems.* * Fu io,' ' la onde,' ' gia era : ' in each of these cases both the syllables are accented, and elision does not suggest itself. We have, then, before us the fact that Tasso quotes with apppoval examples of concurrent vowels, left unelided, in verse, among which a certain number are unquestionably left open. Now, as effects which a poet sanctions in the verse of others he is likely to make use of in his own, it becomes a matter of some interest to enquire, firstly, whether Tasso does himself use concurrent vowels without elision; and, secondly, if he does, how far the usage goes. In the 15,330 verses of the * Gerusalemme ' there are not, I believe, to be found more than twenty-six instances of ' prima facie ' hiatus ; and the vowel combinations they pre- sent are as follow : — Chi h (occurring twice), Tento ella, Cosi, or si alto (3 times), Ma ella (twice), N^ atto, — h (twice), Tre anni, — esce, * In the verse, Io, io vorrei che il vostro alto valore, the words ' io,' ' io ' count for but two syllables ; and it does not seem pos- sible to pass from one to the other without closing the lips. From this movement a ' w ' digammating power results. Now, if the words were pro- nounced as having each two syllables (as in ' vid' io '), would not, also, the ' y '-consonant sound assert itself between the * i ' and the ' o ' of each word ? This line of argument might be carried farther with a view to show that the said consonant powers are both often present in other modern languages be- sides our own. g2 84 petraech's practice. Pivl alto (3 times), Ma esso, — atto, — ecco, — aspre, — empi, — oltre, — odi, Pu\ e piii (3 times). Da ire. Here, then, one remark obviously occurs : if Tasso in- deed deems hiatus to be a means of good effect in verse, it is a means, at any rate, by him used with extreme chariness. Consulting Petrarch, we find him to furnish about fifty- two similar examples ; of which the vowel combinations are, — Chi^, Gi^ era. Te essendo. Cosi (or si) alto (6 times). Tu ora. E di te e di me. — aspre (twice). occhi, Ma^ — or. — anime. Da inde (twice) : N^ altro, — ora, Qui^, — era, — ultimo. Di e notte (twice). Deh ! or. — aspettata. Cosi al lume, Pill alto (3 times), — invidia. Tu hai. — altri. — inconstanza. Ch'io odo, — oltre. Fu io. — era. Faro io (twice). — oso. Artu, e tre Cesari, Sar5 io. E' or. Per6 al. La oltre. — oggi. — i miei. — onde, Fair, — i di miei : and effects of this kind are very common in Dante. It appears, then, that in dealing with exceptional cases of concurrent vowels, Tasso lays down for himself three rules ; namely, 1. When two accented vowels meet, as in ' cosi alto,' ' tentb ella,' elision is to be forborne : 2. Certain unaccented monosyllables (like ' ma,' ' da,' ' ne,' * piu ') are not to be elided before an accented initial vowel ; 3. Nor ' piu,' being accented, in the phrase ' piu e piu,' be- fore an unaccented initial. With one exception,* Petrarch's practice is found to be just the same, as regards these points ; and, further, we find that Once only, ' Lh, onde ' is elided by Petrarch ; La'^onde io passava sol per mio destino. Petrarch's practice. 85 1. He forbears to elide the accented *u' of Artu before an unaccented initial ; 2. Also the unaccented * tu,' and ' fu ' * in the same position : 3. In the phrase * di e notte/ he elides * di/ or not, at pleasure ; but, oftener than otherwise, he does not elide it : 4. In two instances, he leaves ' cosi ' unelided before an unaccented vowel ; while, in several other cases, he adopts the contrary mode, e.g. ' e cosi'^avvien : ' 5. * Te ' is twice by him left unelided before an unac- cented initial ; * The pronoun 'tu' when immediately preceding its verb, *fu' when auxiliary, and ' piii ' when followed by an accented syllable, must each be held unaccented. Now, the vowel-sound in each of these is the same as that of our particle 'to:' hence the Italian pronoun 'tu,' when unaccented, sounds just the same as our said particle. Nevertheless, I cannot produce a single instance in which the former is elided before an accented initial vowel ; while we occasionally do elide our particle in such a position ; thus, A tow'ring eagle to'^all the fowls he seems. The reason, I suggest, is, that though both are unaccented, both are not equally unimportant : hence, the particle, being intrinsically insignificant, is marked by an invariable fugitiveness of sound, which is by no means a characteristic of the pronoun. But be that as it may, such concurrences seem systematically to be avoided by Italian versifiers ; and when found to be unavoidable, are not held to require elision : thus, we see that Petrarch and Tassodo not elide in * piii alto,' etc., nor Petrarch in ' tu ora,' 'fu oso,' nor Dante in ' tu ardi ; ' and yet, I do not observe that any consonant power makes itself felt between the unaccented ' piii ' or ' fu,' and accented initials, as in Piu alto, — atto, — aspre, — oltre — , notwithstanding that betAveen our particle ' to ' and corresponding initial vowels, as in To alter, — act, — ask, — open, there is a quite perceptible sound of the w ; and the same sound, I should say, is perceptible in ' tu ora,' ' tu ardi.' The difference of effect is due to diversity of initial consonant sound in the prior word ; and that this is so will be seen if we substitute an initial ' p ' or ' f ' for the initial ' t ' in 'to alter,' ' to ask,' etc., and we shall find that the w sound is no longer percep- tible. In pronouncing ' few ' or ' pew ' the lips open ; in pronouncing ' to ' they tend to close. The ' u ' vowel-sound is the one of all others which can least easily be elided when it meets with resistance : accordingly, we find that the Italian tongue does not possess a single word, other than monosyllables, ending with an unaccented ' u ; ' nor are the monosyllables tmaccented except Avhen used as proclytics. Moreover, the Latin tongue has not a single short final ' u.' 86 THE MORE SONOROUS THE LESS ELIDIBLE. 6. And thrice, in the same position, the final ' o ' of ' perb : ' 7. The interjection ' ' is four times left unelided when followed by initial accent, and thrice when not followed.* Now, doubtless, in the greater part of these words, the vowels are left open ; but it is not so (as I believe) whenever the concurrent vowel sounds are the same as those from which, in English, the digamma results. To me, at least, it seems that, under similar conditions, similar effects would result in every language. In saying this, however, I wish to be understood as merely stating a theoretical opinion, f As regards the monosyllable * ma,' there is no doubt that Italians do dwell on it (as do also the French on their ' mais ') far more than we ever dwell on our mean-sounding equiva- lent conjunction : but whether this be due to any cause other than mere caprice of elocution, I do not venture to say : again, some vowels are more sonorous than others ; and the final vowels of monosyllables are more sonorous than the same vowels when found at the close of other words ; and the more sonorous a sound is the less easily does it lend itself to elision under the conditions supposed. The fact, at any rate, remains, that Dante, Petrarch, and • There is a diflference to be observed between the interjection * o' or 'oh,' when used as the mere sign of a vocative case, and the same when used to exhort, invoke, or adjure : in the one case it does not take accent, in the other, I should say, it does ; and this seems to be acknowledged by Petrarch when in the lines 0, aspettata in ciel anima bella ! O, invidia nemica di virtude ! 0, inconstanza delle umane cose ! he abstains from eliding before an unaccented initial ; notwithstanding that the general rule of Italian versification is certainly to elide the interjec- tion in such position. Before an accented initial, Petrarch, I believe, always forbears elision. The usage of other poets after his time, so far as my observation goes, seems to be entire avoidance of such concurrences. At least, I do not know of any case in which this interjection is elided before an accented initial. In Latin verse we know the rule is never, under any circumstances, to elide the interjection. f It is a fact here worthy to be noticed that Metastasio, and Alfieri, hardly ever, if at all, allow any concurrences of unelided vowels, such as those "which, sparingly (as we have seen) used by Tasso, less sparingly by Petrarch, are often used by Dante. They seem, indeed, systematically to avoid all combinations which place them under the alternative of eliding, in such cases, or of forbearing to elide. TWO KINDS OF HIATUS. 87 Tasso persistently refuse to elide * ma/ or * da/ or * fu/ or ' tu/ or ' n^/ before an accented initial vowel ; and this can only be because these monosyllables have been felt by them to be, if not unelidible, at least better left unelided, under such conditions. We must allow, then, that there are two kinds of hiatus ; one, in which the dissonance arising from an incomplete sound is aggravated by ai-tificialness of utterance ; another, in which the dissonance exists simply, without any such ag- gravation. The former kind is absolutely rejected in Italian verse ; the latter, if not deemed quite inadmissible, is yet in fact very sparingly admitted. In answer, then, to the question, raised in p. 82, whether a concurrence of accented and undigammated vowels may be allowed in vei*se without elision, I have to say, firstly, that elision never ought to take place between two accented vowels; secondly, that the efiect of it, when only one is accented, is at times unsatisfactory : it would be better, then, to avoid all such concurrences ; * but if that be found im- possible, then the preferable course would be to leave the * Though the term * hiatus ' is generally applied to the gap caused by the concurrence of unelided vowels in separate words, we ought not here to overlook the fact that many languages, more or less, and Latin and Italian as much as any, abound with words in which concurrent vowels are sounded as separate syllables ; as, for instance, in Italian ' glorioso,' ' odioso,' ' ori- ente,' ' soave,' ' fastidi'oso,' ' filial,' * impartial,' ' Religion,' ' desi'ar,' ' maestro,' 'pSese,' 'vxaggio,' ' trionfante,' etc. etc. This treatment is, no doubt, excep- tional ; but still, it always is observed in certain words ; and the number of them is not a few : now, in some such cases, one of the two vowels may, perhaps, be digammated ; in some, neither of them is, or can be : hence it follows that hiatus within words is a thing which does occur occasionally in verse ; and as our ears endure it under such conditions, while, for the most part, they are intolerant of it between words, a question arises how the apparent inconsistency can be explained ? One reason is obvious : as such and such is the right way of pronouncing certain words, they must needs be so pronounced in poetry, or not used at all ; and not to use them would be to render poetry well-nigh impossible : the ear, therefore, recognising the necessity, submits ; and submits with less repugnance, because the effects which result, whether satisfactory or not, are natural effects : but for hiatus between Avords, it has less tolerance ; because, in excuse for them, neither necessity nor naturalness can, in most cases, be pleaded. That interior hiatus is quite satisfactory, I do not say : all I say is, that there is no help for it. Like keyed musical instruments that cannot perfectly be tuned, language itself is an instrument at bestimperfect ; and the utmost to be expected in verse is, that it shall be as perfect as language permits. Now this degree of excellence is found in Italian verse. 88 HYPERMETRICAL SYLLABLES. TWO KINDS. vowels open ; for though the result be a fault of sound, yet when vowels do not easily coalesce, hiatus between them loses a large element of its ofFensiveness ; and if English poets never did anything worse than allow such hiatuses, now and then, after the manner of Petrarch and Tasso, there would not be much to find fault with in their versification. Having now done with the general laws of verse, we have yet to speak of hypermetrical syllables. These are of two kinds ; one occurring in the body of an heroic verse, another at its close : occurring at its close, a weak added syllable makes what is called a double ending, and renders verse endecasyllabic ; and endecasyllabic, in this sense, is, with rare exceptions, as I have said, the heroic verse of three Romance languages : on the other hand, superfluous syllables are absolutely excluded in them from the inner structure of a verse : but with us the case is different ; for there is scarce one among our early dramatists whose lines are free from these ungainly excrescences. The following extracts will suffice for samples of such efiects : * The harbinger to prepare my entertainment — , Thou didst rise goriously, kept'st a glorious course — , His resolution to part with his estate — , With sprigs of eglantine : then a bubbling spring — . Lines like these are essentially licentious ; j- and if they * Several instances of this license occur in ' Comus,' but none in Milton's later poems. f Professor Craik, however, decides differently ; his words are : ' Further, in any place which may be occupied by an unaccented syllable, it is scarcely an irregularity to introduce two or even more such unaccented syllables. The effect may be compared to the prolongation or dispersion of a note in music, by what is called a shake. Of course, such a construction of verse is to be resorted to sparingly, and only on special grounds and occasions : em- ployed habitually, or very frequently, it crowds and encumbers the rhythm, and gives it a quivering and feeble character. But it can in no case be said to be illegitimate, although, in ordinary circumstances, it may have a less agreeable effect in some places of the line than in another.' ('English of Shakespear,' p. 32.) There is no point of resemblance, and nothing, therefore, to compare, be- tween shakes in music and superfluous syllables in verse : regular or irregular such syllables will be deemed, according to the standard we each have in view, SHERIDAN S THEORY OF pass muster, every kind of deformity may be held entitled to the same indulgence : a peculiar rude rhythm of their own such lines may be thought to have ; but they are not heroic verse, nor do they cut a good figure in its company. But here we must distinguish between syllables which, like these, are bona fide hypermetrical, and those which, without being so, are by some treated as such : for some writers hold that syllables which by reason of elision, or synseresis, or slurring, or natural muteness, have (as I and others hold) no effect on the metre, ought distinctly to be sounded, as integral parts of verse, even when the metre is complete without them. Among those who take this view are Thyrwitt, Mr. Steele, Mr. Chapman, Mr. Mitford, Sheridan, and Sir Egerton Brydges. Let it suffice, however, to quote the two last. We have already seen how Sheridan, assuming our heroic metre to consist often of more than ten syllables, gives thirteen to Milton's verse And many a frozen, many a fiery Alp : following out this assumption, he goes on to say, ' I will now produce a couplet of as fine a sound, perhaps, as any in our language, where the former line has fourteen and the latter twelve syllables ; — And many an amorous, many a humorous lay, Which many a bard has chanted many a day.' Then we have Sir Egerton Brydges ; who, in reference to Dr. Johnson's criticism on Milton's verse, says, ' The critic's false principle of our verse continually leads him to blame as faulty what is in truth harmonious : thus, having said that the elision of one vowel before another is contrary to the genius of our language, he is often driven to make this elision by his false rule, as in the line Wisdom to folly as nourishment to wind.' and Professor Craik's standard is different from mine ; but if, while deemed regular, they are yet to be resorted to on special grounds only, one is fain to enquire what the special grounds are. 90 SIR EGERTON BRIDGES. He then takes in hand twenty-two lines of Milton, and scans them in his own way : some of these, as they bear on the question under review, I here produce with Sir Egerton Brydges' scansion : For we have also our evening and our mom — , F5r we | have als5 | our eve|nlng and | our mOrn — , Inhospitably, and kills their infant males — , Inhos|pIta|bly and kills | their Injfant males — , God made thee of choice his own, and of his own, — God made | thee 5f choice | his own |, and of j his own — , Abominable, unutterable, and worse — , Abom|Ina|bl6 unut|t6ra|bl6 and | worse — , Impenetrable, impaled with circling fire, Impen|etra|bl6, im|paled with | circling | fire.* * The scansion of the remaining seventeen lines is of a piece with the above. Sir Egerton Brydges thus states his own theory about Milton's ver- sification : * I believe that Milton's principle was to introduce into his lines every variety of metrical foot which is to be found in the Latin poetry, especially in the lyrics of Horace : such are not merely iambic, but spondee, dactyl, trochee, anapaest ; and that whoever reads his lines as if they were prose, and accents them as the sense would dictate, will find that they faU into one or the other of these feet, often ending, like the Latin, with a half foot : whenever they do not, I doubt not that it arises from a different mode of accenting some word from that which was the usage in Milton's time. If there be any attempt to read Milton's verse as iambics, with a mere occa- sional variation of the trochee, or spondee, they will often sound very lame, instead of being, as they really are, magnificently harmonious.' Those who allege our present mode of pronouncing words to be different from that practised in Milton's time, and use the allegation to support some theory, should, at least, state their reasons : I have already stated mine for disbelieving in any difference between the pronunciation of the two epochs. It is not easy to understand how anyone could take up the theory that Milton versified with Latin feet. He himself says that his metre is English heroic verse, which he describes as consisting of ' apt numbers, fit quantity of syllables, and the sense variously drawn out from one verse to another : ' now, if the ' apt nu.mbers ' here meant be those used by Horace in his lyrics, Milton, doubtless, would have told us so ; for no one, certainly, would find it out by intuition, or any process of reasoning yet recognised among men. However, here we find applied to English verse a system of prosody which modern tongues abhor : for it is vain to talk of Latin feet, unless we accept, with all its consequences, the prosodial S3'stem on which those feet depend. But, apart from that, what metrical effect, I ask, could possibly result from feet of all sorts thus jumbled together without method? When Cicero advises that feet should be so jumbled (' permistos et confusos ') in oratory, he seeks to render it on the one hand rhythmical, on the other free from all semblance of metre. (Orat. Iviii. 195.) But manifestly. Sir Eger- ton Br3'dges has not the faintest notion of what metrical harmony means ; nevertheless, he takes great offence at Dr. Johnson's criticisms of Milton's verse, and marvels at ' his want of ear : ' now Johnson may not always be TWO TESTS APPLIED. 91 Let us now consider the theory put forth by these authors. — When a question arises whether syllables ought, or ought not, to be separately sounded in verse, there are two ways of testing the matter : one is, to insist that the syllables all really have a distinct separate utterance ; another, to in- sert, in place of the questionable syllables, others which must unquestionably be uttered : the latter test I have already applied to a verse quoted by Professor Craik ; and I will now apply them both to the couplet above mentioned, as cited by Sheridan, in support of his theory. This couplet, read, as in my judgment it ought to be read, thus, And many'^an am'rous,many'^a hum'rous lay, Which many'^a bard has chanted many'^a day, consists of two decasyllabic lines, as smooth and natural as any in the ' Deserted Village.' According to Sheridan, it ought to be read thus, 1 2 3 4 5 6 7 8 9 10 11 12 13 14 And many an amorous, many a humorous lay, 1 2345 6 78 9 10 11 12 Which many a bard has chanted many a day ; but let anyone attempt thus to read it, taking care that the whole alleged number of syllables be, each and all, made right, and his ear may not he perfect, but Sir Egerton Brydges is never right save by chance ; and as for ear, he has simply none at all : give him a good line, and his very touch spoils it ; give him a bad one, and that has not fair play ; for he sees not wherein it is faulty, and makes faults in it unconsciously where none exist : thus, out of the twenty -two lines he quotes from Milton, all being by him deemed harmonious, the far greater part are indeed excel- lent, leaving nothing to desire, while some few are defective, more or less, though not in his sense defective : yet notwithstanding the admitted perfec- tion of most, and the alleged perfection of all, there are but two or three of the whole lot which do not come from his hands in a dolefully disfigured plight. Quite forgetting his theory about Milton's principle of versification, he talks of reading verse as though it were prose, and of accenting as the sense dic- tates ; and here he talks well ; for there is but one right way of reading, and sense and prosody should never be at variance ; yet so grossly, in practice, does he pervert prosody and mangle metre, that the best lines, as read by him, have neither the sound of verse, nor of aught else ever heard in lan- guage ; for though lame lines, read rightly, sound well enough as prose, yet the prose he makes of good lines read wrongly is itself found lamer than the lamest verse. Nor do these remarks apply exclusively to Sir Egerton Brydges ; they ^pply to all who adopt systems of versification which cannot be carried out without falsifying the due sound of sj-Uables. Any system, of which that can be said, is stamped, on the face of it, with falsity. 92 DOUBLE ENDINGS IN ENGLISH TERSE. inexorably perceptible to the ear, and not only will the pro- cess cause a painful vocal ej3brt, but the effect will be arti- ficial and grotesque in the extreme. Reduced to mere articulate sound, the effect will be, — Turn turn, | turn turn t6m | turn turn ttim | turn turn turn | turn turn ttim Turn turn | turn turn turn ] turn ttim ] turn turn ] tmn turn turn | ; and reduced to words, it will be, — And numberless beautiful, numberless wonderful lays, Which numberless bards have chanted numberless ways : * and of this, no doubt, the effect is neither artificial nor gro- tesque : and the reason is, that each syllable is here un- avoidably to be sounded ; and each has its natural proper sound : nor is there wanting some sort of rhythm ; | but surely, no one will pretend to call it the rhythm of heroic verse. Though few, I think, will deny that the double endings of Italian, Spanish, and Poi'tuguese verse have an effect far better than the single ones of ours, there are obvious reasons why such endings can be used but to a limited extent in English verse. There is, firstly, the exceeding monosyllabic character of our language, and the great superabundance of accented monosyllables, that is, of verbs, substantives, adjectives, and adverbs : secondly, there is a tendency to accentuate dis- syllables on the final syllable, rather than on the penultimate ; thirdly, the further tendency to throw back the accent, away from the penultimate, in trisyllables, and polysyllables. From these causes, the number of words in our language which are accented on the penultimate is, comparatively, very small. * That Sheridan himself did ever read out this couplet as it must inevit- ably be read if his theory be right, I do not believe ; but if he did not, then the superfluous syllables, claimed by him, would not be sounded ; that is to say, unconsciously, and in spite of himself, he would elide the elidable ones, and cut out the mute. f The words here put together are in fact anapaestic verses, only with five beats instead of four ; being to the regular anapaestic verse what the alexandrine is to the heroic. JOINT WORDS NOT SUITABLE. 93 Nor among those which are so accented, have we many which give, when used at the close of a verse, an effect so good as that given by words of the same class in the Romance tongues : for all such Italian words, and most Spanish and Portuguese, end with a vowel ; while not only do the far greater part of such words in English end with a consonant, but many end with two or three. Again, the final unaccented syllable of English words is often found to be a monosyllabic adjunct, entitled, in its separate state, to accent ; as, for instance, in dissyllables like * birth-right,' ' king-craft,' ' out-cast,' ' strong- hold ; ' and reasons, may be given why the final syllable, of any such word ought not to be used as an hypermetrical ending : * but this state * For though the syllables, in their joint state, may be deprived of accent, they do not, however, lose any part of that significance on which, in their separate state, the title to accent rests. In fact, they are eminently significant, each one being the chief word of a phrase. Such words tacked on, as enclytics, to dependent parts of speech, present a simple grammatical absurdity ; and a further absurdity results from using them as double end- ings ; for these are metrical superfluities ; and reason revolts against allow- ing the chief word of a phrase to pass for a superfluous sound in any sense. Hypermetrical syllables, then, should never draw attention to them- selves : they should either have no meaning of their own, or none which is not dependent on the preceding word. But besides, these adjuncts, deprived of accent, are yet found to claim, by way of compensation, a far fuller utterance than that of other weak final syllables ; all which, indeed, are felt to be best uttered when the ear takes least heed of them. Now, when joint syllables are used in the body of a verse, and are essen- tial to the metre, no ill effect results ; for so long as they be deprived of accent, it is immaterial how they be sounded in other respects ; but if they be used as final syllables of endecasyllabic verse, the eft'ect is bad : for hyper- metrical endings require fugitiveness of sound, and such is not the property of these adjuncts. Indeed, it is questionable whether even the accent may not be transferred to them at pleasure : for my own part, I should not like to pronounce faulty any verse of which the rhythm depended on such a transfer. As regards words like 'peaceful,' 'guiltless,' 'utmost,' 'statesman,' ' stedfast,' ' outrage,' and many more, the case is different : for the final syllables of such words are integral parts or inflexions ; and quite devoid of that grammatical importance, which, at first sight, they may seem to have. Thus, ' peaceful ' is but an adjective which corresponds with ' peace,' and no more presents a complex idea than 'crafty' or 'zealous,' which are adjec- tives corresponding with ' craft ' and ' zeal ; ' ' guiltless ' is the negative of ' guilty,' and synonymous with ' innocent,' only bearing the mark of nega- tion on the final syllable instead of the antepenultimate ; ' utmost ' is the superlative of ' outer,' as ' greatest ' is of ' great ; ' 'statesman ' and ' plough- man ' are simple words, not less than ' warrior ' or ' miller ; ' ' stedfast ' no less so than ' steady,' and in * outrage ' (from ' oltraggio ') the substantive 94 NO GOOD REASON FOR EXCLUSION. of things is never found in the three Komance languages above mentioned ; where every final syllable is an integral part of some single word, in its primary or inflected form. Nevertheless, although, through the peculiarities of our language we are debarred from an uniform use of double endings, and although,' through the same cause, the effect of them in English verse is generally far inferior to that which they have in other tongues, still, I see no reason why they should be excluded, as by many of our poets they are,* from ' rage ' has no part. Accordingly, the final syllables of such words almost invariably drop the primary full sound they have as separate words, for one which better marks their inflexional character : thus, ' full ' becomes ' fle ' (as in * trifle '), ' most ' becomes ' must,' * man ' ' mun,' ' fast ' ' fust,' * rage ' ' redg,' and even the * less ' of ' guiltless ' is not quite the same as ' less ' by itself. Words ending in * like,' as ' warlike,' are an exception ; but though ' warlike ' bears to * war ' the same relation which ' peaceful ' bears to * peace,' the final syllables of these adjectives do not seem to correspond in character: *ful' is clearly an inflexional adjunct; but 'like' seems to assert itself as an adjunctive word. * Thus Thomson excludes them from his ' Seasons,' Akenside from his * Pleasures of Imagination,' Young from his ' Night Thoughts,' and Cowper from his ' Task ; ' yet those of them who are dramatists as well as writers of didactic poetry do not apply this rule of exclusion to their dramas : for Thomson admits them in his * Tancred and Sigismonda,' and Young in his * Zanga.' The reasons given by Dr. Johnson are curious : ' Endecasyllabic lines,' he- says, 'ought not to be admitted into heroic poetry, since the narrow limits of our language allows no other distinction of epic and dramatic measures than is afforded by the liberty of changing at will the terminations of dramatic lines, and bringing them by that relaxation of metrical rigour nearer to prose.' (' Rambler,' No. 88.) Sheridan, too, thus writes on the same subject : ' Lines of this class should seldom be used except by writers of tragedy, whose business it is not to be too curiously solicitous about the melody of their metre, that the dialogue may appear more natural' (' Art of R.,' vol. ii. p. 241) ; and similar ideas are often found cropping up in English literature. Now, such remarks are based on two assumptions, both of which are, in my opinion, false : one is, that epic poetry needs verse of a better quality than dramatic poetry needs ; the other that a verse of eleven syllables is intrinsically inferior to one of ten. Any argument used to excuse laxity of versification in one kind of poetry may as reasonably be urged to excuse it in aU kinds, except, perhaps, some forms of burlesque ; but to change the endings of lines from ten to eleven syllables does not involve ' any relaxation of metrical rigour ;' for eleventh syllables are outside the metre ; nor does it involve any departure from the best models ; for the verse of those languages which furnish the best models is almost invariably endecasyllabic. But the strangest of all fallacies is that verse can be made more natural by making it more like prose. The sole difference between prose and verse is that, both being constructed of natural materials, in one the materials range themselves without method, whereas, in the other, they are arranged OF WHAT DOUBLE ENDINGS CONSIST. 95 our heroic blank verse, or why we should hesitate to use them, side by side with single endings, after the manner of Milton and the dramatists. At the same time, we must distinguish between effects of this kind which are legitimate, and those which are not. As double endings, then, consist of an accented and an unaccented syllable, it is obvious that the latter must either be the final syllable of a word accented on the penultimate, or an unaccented monosyllable which attaches itself in- separably to some accented final; such words are called enclytics ; and of these our language possesses three ; viz. : 1. The personal pronoun, through all its oblique cases, when, not being in antithesis, it follows a verb : * 2. The negative particle * not,' following a verb : 3. The particle of connexion * then.' The following examples show the proper use of these monosyllables as enclytic endings of heroic verse : Nothing of all these evils hath befall 'n me — , Such a discomfit as shall quite despoil him — , To his due time, and providence, I leave thee — , With rueful cry, yet what it was we hear not — , "Where the heart joins not, outward acts defile not — , By that sin fell the angels, how can man, then, The image of his maker, hope to win by it ? — ; according to fixed rules of art : thus, prose is, ' prima facie,' the more natural ; but in all tine arts a something is assumed ; and the thing assumed in poetry is the naturalness of verse ; so that, as regards nature, the two kinds of composition are placed, at starting, on a par : but verse being natural in an assumed sense, cannot possibly be made more natural in another sense without ceasing to be verse : to talk, therefore, of approximating it to prose, is to ignore the postulate on which all poetry rests, and to confuse between two things essentially distinct. * It was said above (p. 10) that the personal pronoun nominative, when immediately following its verb, is (except in antithesis) unaccented. A question, therefore, may arise whether it can be used as a double ending at the close of endeca syllabic verse ? Can it, for instance, be so used in ' where am I ?' ' what say'st thou ? ' ' thus have we,' ' there remain they ' ? In my opinion it never ought : because though inseparably attached in most cases to the antecedent verb, it is not so as an enclytic : for a true enclytic de- pends in sound and sense : thus oblique cases of this word meet both requirements ; its nominative the first only ; while so far is it from meeting the second, that what we here have is, not the pronoun in sense dependent on the verb, but the verb itself dependent on, and directed by, the pronoun. In fact, the reasons which apply against the use of adjunctive words as double endings apply equally in the case before us. 96 FALSE ENCLYTICS: and these next give examples of other monosyllables im- properly used as enclytics in the same position : If in the flow'r and strength of youth, when all men — : this is the only undoubted case I find in Milton : the follow- ing are from Massenger, and others of his time : — The Christians are pursued : he makes his stay Jiere — , Preserve this temple, builded fair as yours ^«— , Which is the end I aim at, being to die too, — 'Tis not the fear of death that makes me sue tMi8 — , Thou family of fools, live like a slave still — , I'll meet you presently : retire av/hile all — , There take this maid: she's at your own dispose now. It would be easy to quote a multitude of similar examples from the pages of Massenger and his cotemporaries. Now, these lines are licentious, because in each of them a word entitled to inalienable accent is treated as enclytic, and used as an eleventh hypermetrical syllable : and how much such words are here out of place is seen at once if, slightly altering some of the lines, we bring the now eleventh syllable into the tenth place, and the now tenth into the ninth place : thus. If in my flow'r of manhood, when all men — , Preserve this fane, fair builded as yours is — , It is not fear that brings me to sue thus — : here, the words before treated as unaccented and hyper- metrical, make just as good endings of decasyllabic verse as any other forcible monosyllable would make ; while the syllables to which they were tacked on, as enclytic, become to them subordinate by operation of a law already pointed out. It was said above that the personal pronoun in its oblique cases after a verb is enclytic unless there be antithesis : thus, in the passage. His empire, and with iron sceptre rule Us here, as with his golden those in heav'n — , the pronoun is properly detached from the verb because of TRUE. 97 the opposition between * us ' and ' those ; ' but in this next example, Oh 1 had his pow'rful destiny ordain 'd * Me some inferior angel — , the pronoun is improperly detached, and the rhythm spoilt in consequence. The following verses of Milton afford instances of the negative particle ' not ' improperly used as a tenth accented syllable : — The anguish of my soul, that suffers not — , To come and play before thee : know'st thou not ? — , I beg and clasp thy knees : bereave me not — : for an enclytic still retains its chai'acter, notwithstanding the intervention of unaccented final syllables, or of other enclytics. I have already pointed out that the monosyllable * then ' is entitled to accent, or not entitled, according as it be an adverb of time or a particle of connexion : but this distinc- tion, though an important one, is seldom observed by our poets ; not even by Milton, who often uses the expletive for a tenth syllable, as in Will covet more : with this advantage, then — , We sunk thus low : the ascent is easy, then — ; and never once, that I am aware of, as an eleventh. The following passages show the proper accentuation of * then ' when used as an adverb of time, At these sad tidings. But no time was then For sad indulgence to their fear or grief — , The happy isle. What strength, what art, can then Suffice, or what evasion bear him safe ? — * One law applies equally to enclytics and to proclytics : the former are attached inseparably to antecedent words, the latter to words follow- ing. Mr. Chapman, by way of showinsc the absurdity (as he thinks) of not sounding the second ' e ' of v;hether ' in the line All judgment whether in heav'n, or earth, or hell, says ironically that ' the author is much indebted to Swift for his authority, which is established by the following lines. And thus fanatic saints though neith'r in Doctrine or discipline our brethren — :' but his sarcasm is not to the point : for Swift's line is not faulty through abscission of the vowel ' e,' but through detachment from its noun-substaa- tive of the preposition 'in.' H 98 RHYME. and the lines next to be quoted show the adverb of time, and the particle of connexion each used as a ninth syllable; a position where accent, though never required, is always ad- missible, in subordination : The deep to shelter us ? This hell then seem'd A shelter to those wounds — , What can we suffer worse ? Is this, then, worst ? Thus sitting, thus consulting, thus in arms ? — Now here, having no metrical inducement one way or the other, what is it that we instinctively dol Beyond doubt, we accent the ' then ' in * this hell then seem'd,' and do not accent it in ' is this, then, worst ? ' Having thus shown how far our method of versification agrees with that used in languages of which the verse sets criticism at defiance, I will now touch on a few other points that ought not to be overlooked in a treatise of this kind. As there are three kinds of endings admissible in verse, so there are three kinds of rhyme ; of which the following are examples : kin'g glo'ry fo'rtify bri'ng, sto'ry, mo'rtifj-, lo've telling e'mulous abo've, swelling, tre'mulous. Rhyme results from identity of vowel-sound in the ac- cented syllables of two or more Words, joined with identity of sound in all subsequent letters or syllables, if such there be. In languages other than ours it often happens that the antecedent consonants are identical also ; so that the rhyming syllables themselves entirely correspond ; as in avenir mains allarmes armatura precorse conviene souvenir, humains, larmes, ventura, corse, diviene: indeed, it is held that the same word may be repeated in rhyme at pleasure, provided it be not repeated in the same sense ; thus the substantive ' porto' (port) may rhyme with ' porto ' (I carry) ; ' versi' (verses) with ' versi ' (thou pourest); PERFECT AND IMPERFECT. 99 and ' campo/ in the sense of ' field,' may rhyme with * campo' in the sense of ' camp.' But entire identity of sound is something more than rhyme ; and our English usage, which avoids such effects, is, to my mind, preferable. Rhymes are perfect, or imperfect : 1. Perfect, when the vowel-sounds exactly correspond, as in love king ago mood debt above, bring, flow, feud, let: 2. Imperfect, when there is between them only a certain resemblance,* or affinity, of sound, as in love reprove thought retum'd grove, love, wrote, mourn'd, arms pest wound war warms, beast, ground, shore. It would be going, doubtless, too far to say that all such as these are inadmissible ; f but thus much may * The only imperfectness of rhyme admitted in Italian is that which results from the difference between the close and open sounds of the vowels 'e' and 'o/ as in ' se'mbra' 'me^mbra,' ' de'ntro' ' ce^ntro,' 'orgo'glio' * fo^glio,' ' co'rto ' ' mo^rto,' ' ro'sso ' ' po^sso,' etc. t Both as regards frequency and sound, it is a question of degree, which must be left to the taste and judgment of the poet. If, however, any poet finds himself often using imperfect rhymes, or ever using one like Flow, then, ye emerald waters, bright and free, And hang aloft, thou rich and purple sky, — (as Mr. Mitford does in a sonnet of his), he would do well to consider whether rhyming be his forte : but as imperfectness of rhyme does not affect the structure of verse, I do not treat it as a fault, provided the rhyming syllables have, each, the tonic accent, and that all subsequent syllables are either integral parts of the rhyming word, or bona fide encly tics. Failing these conditions, the verse, no doubt, is faulty ; though, even then, the fault is not in the bad rhyme, but in the structure from which the bad rhyme results. Walker, in his ' Rhyming Dictionary,' quotes, as rhyming with ac- cented final syllables, a multitude of words, accented on their antepenul- timate. All such are absolutely to be rejected. Thus, he gives to ' bless' 31 true rhymes against 88 false ; to ' us ' 4 true against 62 false ; to ' bent ' 45 true against 104 false ; to ' bate' 33 true against 130 false ; to ' cry ' 33 true against 375 false ; and so with many other endings. Rhymes used in humorous poetry should not, however, be regarded with the strictness due to those used in poetry of a graver kind. Yet, even here, it is a question of degree : thus though we may readily pardon a license like that in Butler's In school divinity as a'ble As he that bight irre'fraga'ble ; h2 100 HEI^OIC COUPLET. TRIPLETS. ALEXANDRINES. at least be said, that the less any of them be used the better.* Our own heroic verse rhymes oftenest in couplets ; that is, by twos and twos continuously throughout a whole poem ; and this simple form seems best suited to the English tongue, which is, of all tongues, perhaps, the one least rich in rhyme. At times, however, a third line is tacked on ; thus making what is called a triplet ; at times, also, a lazy alexandrine is found thrust into the second place : both these variations occur often in Dryden's verse, and occasionally in Pope's : })ut this is all that can be said for them ; for so bad is the eifect that one only wonders how they came ever to be in vogue. The * rima terza ' (which is the form used in Dante's great poem) rhymes continuously by alternate thirds occurring twice ; that is, each desinence occurs three times in regular alternation with some other thrice occurring desinence. The only exceptions are, that the first line of each canto rhymes where a secondary accent is attributed to ' irre'fragable ;' or like the one in Lord Byron's And oh ! ye lords of ladies intellectual, Answer me truly, hav'nt they hen-peck'd you all ? vhere ' all ' is treated as an enclytic ; yet when we come to arhyme like this next, When pulpit drum ecclesia'stic Was beat with fist instead of a' stick ; where accent is attributed to the article ' a,' and the noun-substantive ' stick,' deprived of accent, is tacked on. as an enclytic, to its own article, — here we have an effect abnormal and licentious to a degree that nothing can render excusable. And yet this is one of those rhymes which, according to Lord Monboddo, Butler has ' used with so much success ' ! — (' Orig. and Prog, of L.,' book iii. c. 8.) * In the 868 lines of which Goldsmith's 'Traveller' and * Deserted Vil- lage ' consist, there are only seven containing a rhyme which is not perfect. And as I hare spoken of Goldsmith's poems, I wilfhere add that the ' De- serted Village,' containing 480 lines, has only four faults of versification (that is, less than one fault to every hundred lines), according to the standard adopted in ttiis treatise. In Pope's ' Iliad ' (one of the best versified poems we have) the proportion seems to be about five in a hundred. But the com- parison is hardly fair towards Pope ; for there must needs be far less difficulty in maintaining an unbroken faultlessness through a poem the lines of which count by hundreds, than there is in maintaining it thnmgh a poem the lines of which count by near five times as many thousands : and besides, there are to be found in many parts of Pope's version, some four bundred and odd consecutive lines not showing more faults than those found in the ' Deserted Village.' — I will here add that in Lord Macaulay's 'Vir- ginia,' containing 286 lines, there is no rhyme which is not perfect, and no desinence which does not carrv the tonic accent. STANZAS. THE SONNET. 101 with the third only, and the last but one with the last only.* Stanzas are groups of rhyming verses : the groups contain, each, a complete period, are uniform in arrangement, and de- pendent on each other, as forming part of some poem. The ' rima ottava ' consists of eight heroic lines ; the first six rhyming alternately together ; the two last separately in a couplet. The spenserean stanza has nine lines, eight of them heroic, the last an alexandi'ine. The order of the rhymes is as follows : The 1st line rhymes with the 3rd ; The 2nd with the 4th, 5th, and 7th ; The 6th with the 8th and 9th. These two kinds are used only in long poems. The heroic stanza of four lines, rhyming alternately, is suitable only for short poems of a solemn character, like Gray's ' Elegy.' The Sonnet is a poem complete in itself ; subject to a set of fixed rules, but admitting, within those rules, of many variations. And here (as a standard has to be adopted) I assume, with confidence, that the right standard is the Italian. The fixed rules are : 1. A sonnet contains fourteen heroic lines, neither more nor less : 2. Its lines are arranged in two quatrains, and two tercets ; the former having precedence : * The ' Inferno ' is divided into cantos varying from 115 to 151 lines; and the exceptional rhyming above mentioned is not adopted through ca- price, but with manifest view to a well-considered purpose : for by this contrivance it conies to pass, that though the whole intermediate part of each canto rhymes by twice recurring alternate thirds, each tercet does not, however, rhyme wholly with each succeeding tercet, as will be seen if, starting from the fourth line of each canto, and reckoning thence by sixes, we examine any given six lines, consisting of two complete tercets : — doing this, we shall find that the first desinence of each sestine cori-esponds with one desinence of the foregoing tercet ; so that each sestine is brought into relation with two tercets other than the two of which it is composed ; and thus is avoided the ill-effect which w^ould result from breaking up each canto into a succession of independently-rhyming sestines. 102 FIXED EULES OF SONNET. 3. Each quatrain interrhymes with the Other, each tercet with the other ; but neither quatrain rhymes with a tercet, neither tercet with a quatrain : 4. The first and second lines never rhyme together : 5. The quatrains must be on two rhymes only, equally apportioned; that is, the same desinence will not occur oftener, or less often, than twice in each quatrain : 6. The tercets may be on two rhymes, or on three : 7. Thus, the whole sonnet will not have less than four rhymes, or more than five : 8. The two last lines should never rhyme separately in a couplet. Within these rules,* there are numerous ways of ar- ranging the rhymes. To take the quatrains first : 1. One way is when the 1st line rhymes with the 4th, 5th, and 8th, and the 2nd with the 3rd, 6th, and 7th ; 2. A second, when the 1st line rhymes with the 3rd, * 'La piu difficil maniera che abbia ritaliana poesia, e ad un tratto la pill leggiadra, h il sonetto. Egli e difficile, poiche fra il numero determinato di quattordici versi, senza piii, dee restringere ed abbracciare, con regolata disposiziou di parole, armonia, e chiarezza, una compiuta sentenza ; il che fare ognun conosce quanto sia malagevole; che le piii volte o maggiore minore lunghezza bisognerebbe. Onde, il primo accorgimento del poeta sark, metter la sentenza in tal faccia che si possa o in breve restringere, o ampiamente distendere, senza far torto alia chiarezza, senza bassezza di stile, e senza attastellar vane impertinent! parole. Da quai fregi e pre- rogative, ove vengano interamente osservate, I'altra parte risulta, che h il diletto.' (Zotti, Gr. Ital.) ' In questo letto di Procrusto' (says Metastasio) ' sempre vi si giace a disagio. II nostro Torquato, che ha tanto onorato I'umanitk con suo Gerusalemme, fra la numerosa serie di cento novi e piii sonetti, non ne ha lasciato uno degno del suo nome. L'Omero Ferrarese ne ha due o tre che passano di poco la mediocrita. Nel Petrarca, che ne ha fatta parti- colare professione, non ardirei di vantarne cinque o sei di irreprehensibili. ... In somma, e un componimento al quale gia da molti anni ho creduto prudenza di rinunziare affatto : e tremo per quelli che vi si invilluppano.' (Lettera CCLXIII). Few, perhaps, will share the opinion that Petrarch has not left us more than five or six good sonnets ; yet certain is it, that many a great master of poetry has not excelled in poetry of this special kind ; nor can those who, like Petrarch, have made special profession of it, be said to have done more than excel occasionally. Speaking of Italian sonnets, Sismondi says that the effect of them on our mind results more from the sound than from the thought (Lit. du Mid. de I'Eur. vol. V. p. 401) j such, no doubt, is the case as regards innumerable sonnets : but not so as regards the few which are of the finest quality. VARIATIONS. 103 the 5th, and the 7th, and the 2nd with the 4th, 6th, and 8th. 3. A third, when the 1st line rhymes with the 3rd, 6th, and 8th, and the 2nd with the 4th, 5 th, and 7th ; 4. A foui*th, when the 1st line rhymes with the 4th, 6th, and 7th, and the 2nd with the 3rd, 5th, and 8th ; 5. A fifth, when the 1st line rhymes with the 4th, 6th, and 8th, and the 2nd with the 3rd, 5th, and 7th. Of these arrangements,' four, that is, the first, second, third, and fifth, are used by Petrarch ; the first by far the oftenest ; indeed, out of three hundred, and more, sonnets of his, there are scarcely fifteen not thus arranged.* With regard to the tercets, there are no less than ten arrangements, more or less in use among Italian sonnet- writers of repute. Now, the tercets, as I said before, may have either two or three rhymes : I will first treat of those which have three: 1. The first, and most common, form is, when the 1st line rhymes with the 4th, the 2nd with the 5th, and the 3rd with the 6th ; 2. The second, when the 1st line rhymes with the 5th, the 2nd with the 4th, and the 3rd with the 6th ; 3. The third, when the 1st line rhymes with the 4th, the 2nd with the 6th, and the 3rd with the 5th ; 4. The fourth, when the 1st line rhjnnes with the 5th, the 2nd with the 6th, and the 3rd with the 4th ; 5. The fifth, when the 1st line rhymes with the 6th, the 2nd with the 5th, and the 3rd with the 4th ; 6. The sixth, when the 1st line rhymes with the 6th, the 2nd with the 4th, and the 3rd with the 5th ; . 7. The seventh, when the 1st line rhymes with the 3rd, the 2nd with the 5th, and the 4th with the 6th. f * The fact is worthy to be noticed, as showing Petrarch's preference ; but there is nothing in the form itself to render it intrinsically preferable. t Here, the tercets interrhyme once only; namely, between the 2nd and 5th lines : but notwithstanding this imperfectness of correspondence, the form is well accredited ; being used, several times by Giovan Battista 104 ENGLISH SONNETS. Of tercets with two rhymes there are three arrangements : 8. The first, when tTie lines rhyme alternately ; 9. The second, when the 1st line rhymes with the 3rd, 4th, and 6th, and the 2nd with the 5th ; * 10. The third, when the 1st line rhymes with the 5th and 6 th, and the 2nd with the 3rd and 4th. But this last-mentioned form, though used thrice by Petrarch, is not used, so far as I observe, by other Italians, f It will be seen, then, that the principle which rules the structure of a sonnet is correspondence of rhymes between quatrain and quati-ain, between tercet and tercet. Now, the far greater part of our English poems, called sonnets, are constructed without any regard to this principle.^ Zappi, and others of repute. Among Zappi's sonnets, where we find it, I will mention that beginning, ' Tornami a mente,' and another (much ¥ raised by Muratori), beginning, ' Amo Leusippe.' (See ' Delia Perfetta 'oesia Italiana,' vol. x.) * Used in Petrarch's 66th sonnet, and in the 35th of Giovanni della Casa, beginning, ' La bella Greca.' t Hence the general rules essential to be observed in a sonnet may safely receive two additions ; namely, that the same desinence should never occur three times consecutively ; and that the tAvo last lines should never, under any circumstances, rhyme together. Petrarch's three sonnets, above mentioned, offend on both these points ; and not without good cause, there- fore, has the form in question been discarded. X Three quatrains, rhyming apart, followed by a couplet, was the favourite form of our early so-called sonnet-writers. This is the form used by Shakespear, and several of his cotemporaries : it has also been much used in later times : by Bowles, for instance, by Southey, and by Lamb. The. rhyming of Spencer's 'sonnets,' down to the 9th 'line, is exactly that of the spenserean stanza ; and altogether it is as follows : 1,3, 2, 4, 5, 7, 6, 8, 9, 11, 10, 12, 13, 14. The sonnets of Sir Philip Sydney, of Barnes, and (with rare exceptions) of Drummond, are formed on the Italian model, down to the 12th line ; but they all have the defect of ending with an isolated couplet. At least, this is the ca;se in all the specimens given by Mr. Dyce in his collection. (See ' Specimens of Eng. Sonnets,' selected by Rev, Alex, Dyce, 1883.) Mr, Hallam is mistaken when he says (' Lit, of Eur,,' vol, iii, chap, v,) that Milton's sonnets frequently deviate from the best Italian structure : there are only two of which this can be said, namely, the twelfth and the sixteenth. The fault of the twelfth is that the same desinence occurs seven times, namely, on the 2nd, 3rd, 6th, 7th, 10th, 11th, and 13th lines : that the quatrains and tercets interrhyme throughout, and the sonnet is on three rhymes only. ANAP^STIC A'ERSE. 105 Anapaestic verse, of which we have now to treat, seeks effects through variation in the number of syllables, within a fixed number of beats ; thus differing essentially from iambic verse, which seeks (as we have seen) effects through variation in the number of beats, within a fixed number of syllables. The more common form has four beats, and syllables varying from twelve to nine : * thus, 12. And the clans of Culloden are scatter 'd in flight — , 11. They rally, they bleed, for their kingdom and crown—, 10. crested Lochiel, the peerless in might — , 9. Lochiel, Lochiel, beware of the day. — A verse of twelve syllables consists of four accentual anapaests ; or, to speak more exactly, of four forcible syllables, each preceded by two weak ; f and so far as, in other lines, the syllables are less than twelve, the metre, departing from the primary anapaestic model, partakes of those forms which are adopted in the heroic. Thus, in the line Lochiel, Lochiel, beware of the day, — the first six syllables are accentual iambi, forming an heroic hemistich ; nevertheless, the verse is anapaestic, because the last three syllables are such. A verse of ten syllables may also begin with a single accented syllable, as in Once I on a time | as old stories rehearse ; or with one which naturally phrases with a succeeding un- The fault of the sixteenth is that it ends with a separately rhyming couplet. * Sheridan says (' Art. of R.,' vol. ii, p. 317) that this kind of verse should always consist of four entire feet, containing three syllables each. But a succession of verses thus constructed is soon found to have an unsupport- able monotony : frequent departure, therefore, from the primary model is essential to good effect in this metre as in others. t It has already been shown that there are more ways than one of phrasing most lines': in fact, it is rare to find a line, like the fourth above quoted, which can be phrased but in one way. There is no necessity, there- fore, that any one phrase of a perfect anapaestic verse, as recited, should be of itself an accentual anapaest : the sole thing necessary is that there should be two weak syllables, or two without metrical accent, preceding a strong one metrically accented : but whether the two be ranged, or not, in the same phrase with the strong one, is immaterial. 106 ANAPiESTIC VERSE. accented syllable, so as to make with it an accentual trochee e.g.— Give me ] the man | in whose heart is no guile — . The verse may either rhyme in couplets, or alternately. Campbell's 'Lochiel,' above quoted, and Canning's 'Pilot that weather'd the storm,' are examples of each mode. Another form, versified on the same principle, has three beats, with syllables varying from nine to seven. This form, which always rhymes alternately, is best suited to pastoral ditties, like those of Shenstone, My banks they are furnish'd with bees, Whose murmurs invite one to sleep ; My grottoes are shaded with trees, And my hills are white over with sheep : it may also be used, with good effect, in epigrammatic pieces, provided the movement does not extend beyond eight lines. Lord Byron's epigram, beginning When coals to Newcastle are carried, is a case in point. Lines of this kind, having three beats, assort well with others having four ; as in Moore's poem, beginning Believe me, if all those endearing young charms : but the movement cannot well be sustained beyond two stanzas containing, each, eight lines.* A thii'd form of anapaestic verse has two beats, with syllables varying from six to five ; but this form is never used except in stanzas of six lines, the third and sixth of which, rhyming together, have three beats ; and the stanza is found to be suitable only for humorous and epigrammatic pieces. * Wolfe's famous ' Elegy ' is fundamentally of the same metre ; with this ditference, that whereas in Moore's stanzas, and in most like them, the iambic variations are generall}'- confined to the opening of a verse, or, at least, to the first hemistich, here they occm- in every part ; e.g. By the strugg|ling moon] beam's mis|ty light, And the Ian j tern dim | ly b ur | ning. ENGLISH HEXAMETER. 107 Some notice* is now due to the metre called * English Hexameter/ used fii'st by Southey,f about half a century ago. As English syllables are not subject, like the Latin, to a quantitive prosody, the only way in which there is even a prima facie possibility of our having hexameter verse at all resembling the ancient, is by assuming accented and un- accented syllables to correspond, for the purpose, with long and short ones : but this assumption does not avail in practice ; for the Latin heroic verse consists, we know, of dactyls and spondees intermixed, and of no other feet ; and the ancient prosody permits words to have two, three, four, or five consecutive long syllables, J available for spondees, and each claiming stress (if the metrical efiects due are to be marked) in verse ; and numberless syllables, not otherwise long, become so by position ; whereas our words have never more than one syllable on which stress can be laid ; and ac- cent by position is a thing unknown to us. Unless, therefore, all the first five feet be accentual dactyls, it is physically impossible to construct hexameter verse, on the ancient model, with English words. § In Southey's hexametei*s, as in Yirgil's, the number Oi syllables varies from seventeen to thirteen ; and the first and fourth of the last five are always forcible, the rest of them weak : but in no other respect is there any conformity be- tween the two metres.] Nor does the modern hexameter differ less from the * For an account of other metres not here noticed the reader is referred to Latham's ' English,' part v., and to Guest's * English Rhythms.' f In his ' Vision of Judgment.' The claim put forth by Southey to have invented the metre in which this poem is composed has been made a matter of reproach against him, on the ground that Sir Philip Sydney and others had long before attempted to compose English hexameters on the Latin model : but the principle on which the hexameters of these authors are built we find to be quite different from that adopted by Southey. X The first word of the verse, Fortunatorum nemorum sedesque beatas, has five consecutive long syllables. § That is, without falsifying language : if once we allow ourselves to do that, there is hardly any combination of words which may not be called verse. 11 The hexameters of Sir Philip Sydney are constructed on the principle of subjecting English words to the rules of Latin prosody : this cannot be done without a systematic falsification of our accent ; and we see the eflfects in the lines following : — 108 southey's hexameters, what they are. ancient when the latter is read, as we read it, by accent : nay, between the two, as thus compared, there is found to be, on their respective merits, a further difference, which is material ; for the one has a distinctive rhythm throughout ; while the other has no rhythm which can at all claim to be called distinctive.* If we take a score of Southey's lines, and analyse them, they are found to be decidedly anapaestic in character ; differ- ing from the ordinary verse so called in little | but in this, that they have more syllables and more beats. Syllables they have varying from seventeen to thirteen, and beats varying from six to five ; J and as in the ordinary That t5 mf | advance |ment their ] wisdoms | have m6 S] based] — , Well mSy & | pastor | plain ; biit S.|]as! his | plaints bS nSt | esteem'd J — , Oppress'd | with ruinjous con|ceits hf thS | aid Of Sn | outcry — : Southey, on the other hand, does not so much as profess to form his verse on, the ancient model: 'you try the measure,' he says, 'by the Greek and Latin prosody ; you might as well try it by the laws of Solon, or the Twelve Tables ' (Reply to Rev. G. Tilbrook) : all he professes is to give an imitation so far as the genius of our language permits : in other words, so far as is practicable without departing from the proper pronunciation of English. It is true that some of Sir Philip Sydney's lines, read naturally, have a rhythm similar to that found in Southey's ; but the effects, thus observ- able," result from a treatment which, though quite right, was not intended by the author. * In Latin hexameter, thus read, we often, no doubt, find the ' disjecta membra ' of some modem metre ; and, at times, complete heroic, or even alexandrine, verses : thus, the first line of the Mneid only wants an ac- cent on ' qui ' to render it a good heroic verse, up to that word ; the second is entirely heroic, up to ' Lavinia ; ' and, completed, makes an alexandrine. But verses furnishing such effects are comparatively few : in the far greater number, it seems impossible to detect anj^thing like a distinctive rhythm. As regards the ancient metres, there is one fact, which, being remarkable, may here be noticed ; though, for my own part, I am quite unable to account for it : hexameter verse has no clear metrical effect, except when read ac- cording to quantity ; sapphic verse, on the contrary, has none, except when read according to accent, notwithstanding that, equally with the hexameter, it is composed of quantitive materials. t The poem, it is true, contains some forms of rhj'thm not usually found in anapaestic verse : but these may be shewn not to conflict with the anapaestic character of the movement. 3; Southey says that each of his hexameters has six feet : but here he is quite mistaken. Feet are divisions of verse, marked by beats of time ; and beats of time depend on forcible syllables occurring in certain positions: now, a large proportion (about thirty-five per cent.) of Southey's hexameters have only five forcible syllables. The proportion of Latin hexameters, which, read by accent, have but five beats, is more than eighty per cent., and some have but four. NOT SO BAD AS SOME SAT. 109 anapaestic verse, so in this, the heroic element asserts itself in proportion as the numher of syllables varies between the maximum and minimum permitted. Although, therefore, this verse has but a very slight re- semblance to the ancient hexameter, and has no claim to be called hexameter, save in the sense that, without necessarily having, it often has, six beats ; * still, I quite disagree with those critics who, ignoring all metrical effect in it, pronounce two-thii'ds of each line to be no better than *a rumbling irregular piece of prose : ' | on the contrary, I find in these lines not only a clear metrical effect, but one by no means to be despised. A good deal has been said, by ancient and modern authors, about correspondence in verse between sound and sense : by which is meant not a mere general suitableness of style and diction, but an actual resemblance between the rhythm of verse, and the thing or things described. Such resemblances are in truth only possible when the thing to be described is sound caused by some movement ; and the value of them has been so well appraised by Dr. Johnson, :[: that not much remains to be added on the subject. I will, however, give two examples : premising that any argument which is good for one language on this point is good equally on it for all others. The following line of Homer, — Adrts CTretro ireSovde Ki/A.iv5eT0 \aai avat^rjs — , and the following of Virgil, Quadrupedante putrem sonitu quatit ungula campum — , are often quoted as fine specimens of (what is called) ' word- painting.' Now, both these lines describe sound in connex- * As the ordinary anapaestic verse seeks effects through variation in the number of syllables, within a fixed number of beats, Southey's ana- paestic hexameter must be admitted to suffer detriment in so far as it departs from this principle. f See ' Edinburgh Kev.,' vol. xxxv. p. 427. — These words, applied by the reviewer to Southey's hexameters, are, in my opinion, far more applic- able to the first four feet of Latin hexameter, read bv accent. X ' Rambler,' Nos. 92, 96, and ' Life of Pope.' no COREESPONDENCE BETWEEN SOUND AND SENSE ion with movement : the first, that of a heavy stone rolling precipitately down a hill-side ; the second, that of hoofs tramping, at a pace not stated, over a sun-burnt plain : but though both describe some kind of movement, they describe, however, very different kinds ; and yet both have exactly the same rhythm. If, then, the first line is to be admired for its pictorial property, the second ought, on the same principle, to be blamed for its deficiency in that respect : or if the same rhythm be deemed equally suitable to describe both move- ments, then the alleged correspondence between sound and sense cannot, I should say, be very striking in either case. And how faint, indeed, the resemblance at best is, and must be, becomes evident if we examine any of the two things thus compared : the march, for instance, of poetic feet is slow, measiu-ed, and stately ; whereas stones descend with a rapid, irregular, and impetuous rush. But if a sequence of five dactyls * be still deemed to ha,ve * In a critical notice of Mr. Mitford's * Harmony of Language,' a re- viewer, writing early in this century, endeavoured to' show that the ancient hexameter was based on accent no less than on quantity, and that we have been taught to read Homer's above-quoted verse with a wrong cadence ; the cadence being, however, in this case, none other than such as gives five quantitive dactyls to the verse. ' We beg,' he adds, ' our readers who have probably seen a stone bounding down the steep pitch of a hill, rolling along the slope, and striking at last against some obstacle below, to read the verse according to its real accents, that is, enetTa and ireSofSe like the English word " cruelty," and the last syllable of ai/aiSil}? like the English " dace," and they will find a remarkable instance of what has been called imitative harmony.' ' If it be said tkat in so reading the verse the quantities are falsified, we must ask whether the following Latin line be false in quantity ; which, if the two Jast words be spoken together, will be similar to it: Inde ru'ens per a'gros nemoro'saque te'squafugit su's? ' ('Edinburgh Rev.,' vol. vi. p. 371.) I have quoted this passage as it well serves to show how vain is the attempt to combine things between which there is an irreconcilable incom- patibility. The reviewer starts by begging two material questions ; one being, that the customary marks over Greek syllables denote Greek acute accent; another, that the ancient acute accent was stress, like the modern tonic. Now, both these points have been keenly questioned by good scholars ; (among whom, as regards the first, may be mentioned Isaac Vossius, as regards the second, Dr. Foster) ; and both remain unsettled to this day : so that here we find specifically ascribed to Greek verse a something, of which no one knows, for certain, either what its place, or what its nature, was. SOUND AND SENSE. Ill some special suitableness for depicting agitated movement, beyond doubt, there must be in it a proportionate un- suitableness for depicting a quite opposite state of things, as in Panditur interea domus omnipotentis Olympi — , or Turn Zephyri posuere : premit placida aequora pontus : yet we never hear such lines condemned as presenting images of perfect quietude. How many and how manifold are the conditions favour- able to versifiers which, existing in the Italian tongue, do not exist in ours, must be evident, at a glance, to anyone among us who has a fair knowledge of Italian poetry : nor is it only in those respects which concern the harmonic structure of verse that we are at a disadvantage, it is also as regards those which concern forms of speech, and the grammatical structure of sentences; for ambiguity, and inelegance of diction. However, we are invited to pronounce eireira and neSovSe in such a way as to make of each word an accentual dactyl {-^'•^), instead of what, taken by itself, each is, a quantitive amphibrach (^ -^) ; and in the verse, thus read, we are further invited to discern a remarkable instance of imitative harmony. Imitation there is, so far as all sounds caused by irregular movement bear some resemblance, on the score of irregidarity, to each other : but are the sounds here harmonious ? 1 cannot, for my own part, imagine any- thing having less pretension to be so called. Nor again, do I see that it makes the least difference, as regards the rhythm, whether the final syllable of avaiS^? be pronounced with a vowel- sound, as in ' dace,' or without one, as in ' ease : ' besides, if this line is to be read according to the marks in one part, why not on the same plan in another ? if ejretra and neSovSe are to be changed, because of them, from quantitive ampliibrachs to accentual dactyls, why is ivatS^? to retain its quantitive pronunciation, instead of being taken, on the same principle, as an accentual anapaest (^ ^ -) ? Now, that the quantities are falsified in eneira and rreSouSe, by reading these words in the way suggested, seems to me self-evident ; and yet, to prove the contrary, the reviewer points to a Latin verse in which ' ruens ' and ' agros ' are read accented on their penultimates, and asks whether falsification of quantity results from such treatment '? Certainly, I reply, it does ; seeing that these words, which are quantitive iambi, are here turned into accentual trochees. Then, as regards 'fugit,' on what principle are we to treat this as a quantitive word, if * ruens ' and ' agros ' are to be treated, in the same breath, as accentual ? Once for all, there are two principles of versification, essentially distinct from one another ; and no one who attempts to amalgamate them can help getting entangled in a hopeless muddle of inconsistencies. 112 DIFFICULTIES OF ENGLISH VERSE. in verse, are scarcely less offensive than faults of sound ; and there is not, I should think, under the sun a language in which the difficulty of avoiding these defects is greater than it is in English.* From these causes, it may safely be asserted that with equal imagination on both sides, equal mastery of language, and equal aptitude for numbers, an Italian would with ease compose a thousand good verses in the time taken by an Englishman to compose a hundred such. An English poet, then, may well pause to ponder on the special difficulties of his craft ; because if, ignoring them, he attempts too much, they cannot afterwards be accepted in excuse for failure. Besides, he needs to be on guard against fallacies which, once adopted, make failure certain from the outset. If he thinks, for instance, to make verse more natural than its nature admits, and pleads that object in excuse for using a slovenly versification, the case is hopeless ; for where wrong is done through mere infirmity there is hope ; but we despair of people who do wrong on principle Good poetry needs good verse ; and verse cannot be good if it be not at least free from fault ; and to show what fault is, and how verse may be kept free therefrom, is the main * The fact of our adjectives, participles, definite article, and possessive pronouns ' mine ' and ' thine,' having no inflexions of gender or number ; our possessive pronouns ' his ' and ' her ' none of number ; our substantives, indefinite article, and pronoun ' it,' none of gender ; and, with trifling ex- ceptions, our verbs none of person or number ; — all this is a plentiful source of ambiguity and inelegance. Another cause of inelegance is the multiplication of mean particles, and their ever-recurring tendency to repeat themselves in the same clause or sentence. This tendency, remarkable in many such words, is specially so in the particle ' to,' which, besides being the sign of a case, is also one of a mood, and is further used to express different relations in that mood : thus, it occurs no fewer than six times in the following sentence which is of a type quite common :- ' Jones replied to me that we ought to Avrite to BroAvn to beg him to return to England.' Now, so much more graceful is the struc- ture of other languages, that in Latin, Italian, or French, the same thing would be expressed without any repetition whatever. A single repetition of this particle in verse, as in the line AAvait the morning beam to give to li<;ht, (Thorns. S. Sp. 218) is not without a certain offensiveness : and that such eifects have been sys- tematically eseheweift in English verse is hardly to bs doubted, seeing how very seldom they occur in the pages of our more careful poets. CONCLUSION. 113 object of this treatise : if its tenour seems discouraging, the answer is, that it will discourage none who are competent to succeed ; and competent are those only who combine an innate aptitude for numbers with prompt unhesitating obedience to elementary laws : success, indeed, is shown by the result ; but failure may be augured at an early stage of the process ; for if an artist chafes under the constraint needed to make his wares just tolerable, he is not likely to succeed in ever making them much more than that. LONDON : PRINTED BY 8P0TTISW00DE AND CO., NKW-STEEET SQtJARB AND PARLIAMENT STREET GENERAL LISTS OF NEW WORKS PUBLISHED BY Messrs. LONGMANS, GEEEN & CO. PATERNOSTER ROW, LONDON. HISTORY, POLITICS, HISTORICAL MEMOIRS &c. .Vrmitage's Childhood of the English Nation. Fcp. 8vo. 2s. 6d. ^Vmold's Lectures on Modem History, 8vo. 7*. 6d. Buckle's History of Civilisation. 3 vols, crown 8vo. 24s. Chesney's Indian Polity. 8vo. 21*. — Waterloo Lectures. 8vo. 10s, 6d. Cox's General History of Greece. Crown 8vo. 7s. 6d. — History of Greece. Vols, I & II. 8vo. 36s. — Mythology of the Aryan Nations. 2 vols. 8vo. 28s. — Tales of Ancient Greece. Crown 8vo. 6s. Epochs of Ancient History : — Beesly's Gracchi, Marius, and Sulla, 2s, 6d. Capes's Age of the Antonines, 2s. 6d. — Early Roman Empire, 2s, 6d. Cox's Athenian Empire, 2s, 6d. — Greeks and Pei-sians, 2s. 6d. Curteis's Rise of the Macedonian Empire, 2s. 6d. Inhe's Rome to its Capture by the Gauls, 2s. 6d. Merivale's Roman Triumvirates, 2s. 6d. Sankey's Spartan and Theban Supremacies, 2s. Qd. Epochs of Modern History : — Church's Beginning of the Middle Ages, 2s. 6d. Cox's Crusades, 2s, 6d. Creighton's Age of Elizabeth, 2s, Gd. Gairdner's Houses of Lancaster and York, 2s. Gd. Gardiner's Puritan Revolution, 2s. 6d. — ThMy Years' War, 2s. 6d. Hale's Fall of the Stuarts, 2s. 6d. Johnson's Normans in Europe, 2s. 6d. London, LONGMANS & CO. General Lists of New Works. Epoclis of Modern History :— continued. Ludlow's "War of American Independence, 2s. 6d, Morris's Age of Queen Anne, 2s. Gd. Seebohm's Protestant Bevolution, 2s. 6d. Stubbs's Early Plantagenets, 2s. 6d. Warbuj-ton's Edward III., 2s. Gd. Fronde's English in Ireland in the 18th Century, 3 vols. 8vo, iSs. — History of England. 12 vols. Svo. £8. 18^. 12 vols, crown 8vo. 72s. Gardiner's England under Buckingham and Charles I., 1624-1628. 2 vols. Svo. 24«. — Personal Government of Charles I., 1628-1637. 2 vols. Svo. 24«. GreviUe's Journal of the Reigns of George IV. & William IV. 3 vols. Svo. 36s. Howorth's History of the Mongols. Vol. I. Royal Svo. 28*. Ihne's History of Rome. 3 vols. Svo. 45s. * Lecky's History of England. Vols. I. & II., 1700-1760. Svo. 36*. — — — European Morals. 2 vols, crown Svo. 16*. — Spirit of Rationalism in Europe. 2 vols, crown Svo. 16s. Lewes's History of Philosophy. 2 vols. Svo. 325. Longman's Lectures on the History of England. Svo. 15s. — Life and Times of Edward III. 2 vols. Svo. 28s. Macaulay's Complete Works. S vols. Svo. £5. 5s. — History of England :— Student's Edition. 2 vols. or. Svo. 12*. I Cabinet Edition. S vols, post Svo. 48*. People's Edition. 4 vols. cr. Svo. 16*. | Library Edition. 5 vols. Svo. £4. Macaulay's Critical and Historical Essays. Cheap Edition. Crown Svo. 3*. 6d. Cabinet Edition. 4 vols, post Svo. 24*. I Library Edition. 3 vols. Svo. 36*. People's Edition. 2 vols. cr. Svo. 8*. | Student's Edition. 1 vol. cr. Svo. 6*. May's Constitutional History of England. 3 vols, crown Svo. IS*. — Democracy in Europe. 2 vols. Svo. 32*. Merivale's FaU of the Roman Republic. 12mo. 7*. Gd. — Greneral History of Rome, b.c. 753— a.d. 476. Crown Svo, 7*. 6d. — History of the Romans under the Empire. 8 vols, post Svo. 48*. Osborn's Islam under the Arabs. Svo. 12*. Prothero's Life of Simon de Montfort. Crown Svo. 9*. Rawlinson's Seventh Great Oriental Monarchy — The Sassanians. Svo. 28*. — Sixth Oriental Monarchy— Parthia. Svo. 16*. Seebohm's Oxford Reformers — Colet, Erasmus, & More, Svo. 14*. Sewell's Popiilar History of France. Crown Svo. 7*. 6d. Short's History of tlie Church of England. Crown Svo. 7*. 6d. Taylor's Manual of the History of India. Crown Svo. 7*. 6d. Todd's Parliamentary Government in England. 2 vols. Svo. 37*. Trench's Realiti^ of Irish Life. Crown Svo. 2*. Gd. Walpole's History of England. Vols. I. & II. Svo. (Jn October.) Wyatt's History of Prussia, a.d. 700 to a.d. 1525. 2 vols. Svo. 36*. BIOGRAPHICAL WORKS. Burke's Vicissitudes of Families. 2 vols, crown Svo. 21*. Gates's Dictionary of Greneral Biography. Medium Svo, 25*. Gleig's Life of the Duke of Wellington. Crown Svo. 5*. London. LONGMANS & CO. Jerrold's Life of Napoleon III, Vols. I. to III. 8vo. price ISs. each. Jones's Life of Admiral Frobisher. Crown 8vo. 6*. Lecky's Leaders of Public Opinion in Ireland. Crown 8vo. 7s. 6d. Life (The) of Su: William Fairbairn. Crown Svo. 2*. 6d. Life (The) of Bishop Frampton. Crown Syo. 10^. 6d. Life (The) and Letters of Lord Macaulay. By his Nephew, G. Otto Trerelyan, M.P. Cabinet Edition, 2 vols, post Svo. I2s. Library Edition, 2 vols. Svo. 36*. Marshman's Memoirs of Havelock. Crown Svo. 3s. 6d. Memorials of Charlotte Williams-Wynn. Crown Svo. 10s. 6d. Mendelssohn's Letters. Translated by Lady "Wallace. 2 vols. cr. Svo. 5s. each. Mill's (John Stuart) Autobiography. Svo. 7s. 6d. Nohl's Life of Mozart. Translated by Lady Wallace. 2 vols, crown Svo. 21s. Pattison's Life of Casaubon. Svo. 18s. Spedding's Letters and life of Francis Bacon. 7 vols. Svo. £4. 4s. Stephen's Itesays in Ecclesiastical Biography. Crown Svo. 7s. 6d. Stigand's Life, Works, &c. of Heinrich Heine. 2 vols. Svo. 28s. JSimmem's Life and Philosophy of Schopenhauer. Post Svo. 7s. Gd. — — — Works of Lessing. Crown Svo. 10s. Gd. CRITICISM, PHILOSOPHY, POLITY &c. Amos's View of the Science of Jurisprudence. Svo. 18s. — Primer of the English Constitution. Crown Svo. 6s. Arnold's Manual of EngUsh Literature. Crown Svo. 7s. Qd. Bacon's Essays, with Annotations by Whately. Svo. 10s. 6d. — Works, edited by Spedding. 7 vols. Svo. 73s. Gd. Bain's Logic, Deductive and Inductive. Crown Svo. 10s. 6d. Part I. Deduction, 4s. | Part II. Induction, 6s. 6d. Blackley's Grerman and English Dictionary, Post Svo. 7s. 6d. BoUand & Lang's Aristotle's Politics. Crown Svo. 7s. 6d. BuUinger's Lexicon and Concordance to the New Testament. Medium Svo. 30s. Comte's System of Positive Polity, or Treatise upon Sociology, translated :— Vol. I. General View of Positivism and its Introductory Principles. Svo. 21s. Vol. II. Social Statics, or the Abstract Laws of Human Order. 14 s. Vol. III. Social Dynamics, or General Laws of Human Progress. 21s. Vol. IV. Theory of the Future of Man ; with Early Essays. 24s. Congreve's Politics of Aristotle ; Greek Text, English Notes. Svo. 18s. Contanseau's Practical French & English Dictionary. Post Svo. 7s. 6d. — Pocket French and English Dictionary. Square 18mo. 3s. Gd. DoweU's Sketch of Taxes in England. Vol. I. to 1642. Svo. 10s. Gd. Farrar's Language and Languages. Crown Svo. 6s. Finlason on the New Judicial System. Crown Svo. 10s. Gd. Geffcken on Church and State, translated by E. F. Taylor. 2 vols. Svo. 42s. Grant's Ethics of Aristotle, Greek Text, English Notes. 2 vols. Svo. 32s. Kalisch's Historical and Critical Commentary on the Old Testament ; with a New Translation. Vol. I. Genesis, Svo. 18s. or adapted for the General Reader, 12s. Vol. II. Exodus, 15s. or adapted for the General Reader, 12s. Vol. III. Leviticus, Part I. 15s. or adapted for the General Reader, 8s. Vol. IV. Leviticus, Part II. 16s. or adapted for the General Reader, 8s. London, LONGMANS & CO. General Lists of New Works. Latham's Handbook of the English Language. Crown 8vo. 6*. — English Dictionary. 1 vol. medium 8vo. 24«. 4 vols. 4to. £7. Lewis on Authority in Matters of Opinion. 8vo. lis. Liddell & Scott's Greek-English Lexicon. Crown 4to. 36^. — — — Abridged Greek-English Lexicon. Square 12mo. 7^. 6d. Longman's Pocket German and English Dictionary. ISmo. 5s. Macaulay's Speeches corrected by Himself. Crown Svo. 3«. Gd. Macleod's Economical Philosophy. Vol. I. Svo. I5s. Vol. II. Part I. 12^. Mill on Representative Government. Crown Svo. 2s. — — Liberty. Post Svo. 7s. 6d. Crown Svo. Is. id. Mill's Dissertations and Discussions. 4 vols. Svo. 46*. Gd. — Essays on Unsettled Questions of Political Economy. Svo. 6*. Gd. — Examination of Hamilton's Philosophy. Svo. IGs. — Logic, Ratiocinative and Inductive. 2 vols. Svo. 25*. — Principles of Political Economy. 2 vols. S70. 30*. 1 vol. cr. Svo. 5.?. — Utilitarianism. Svo. 6*. MHUer's (Max) Lectures on the Science of Language. 2 vols, crown Svo. 165. Rich's Dictionary of Roman and Greek Antiquities. Crown Svo. 7*. Gd. Roget's Thesaurus of English Words and Phrases. Crown Svo. 10*. Gd. Sandars's Institutes of Justinian, with English Notes. Svo. IS*. Swinbourne's Picture Logic. Post Svo. 5*. Thomson's Outline of Necessary Laws of Thought. Crown Svo. 6*. TocqueviUe's Democracy in America, translated by Reeve. 2 vols, crown Svo, 16.?. Twiss's Law of Nation?, Svo. in Time of Peace, 12*. in Time of War, 21*. Whately's Elements of Logic. Svo. 10*. 6d. Crown Svo. 4*. Gd. — — — Rhetoric. Svo. 10*. Gd. Crown Svo. 4*. Gd. — English Synonymes. Fcp. Svo. 3*. White & Riddle's Large Latin-English Dictionary. 4to. 28*. White's College Latin-English Dictionary. Medium Svo. 15*. — Junior Student's Complete Latin-English and English-Latin Dictionary. Square 12mo. 12*. r, 4- 1 f The English-Latin Dictionary, 5*. Gd. beparateiy | j,^^ Latin-English Dictionary, 7*. Gd. White's Middle-Class Latin-English Dictionary. Fcp. Svo. 3*. WiUiams's Nicomachean Ethics of Aristotle translated. Crown Svo. 7*. Gd Yonge's Abridged English-Greek Lexicon. Square 12mo. 8*. Gd. | — Large English-Greek Lexicon. 4to. 21*. I Zeller's Socrates and the Socratic Schools. Crown Svo. 10*. Gd. — Stoics, Epicm-eans, and Sceptics. Crown Svo. 14*. — Plato and the Older Academy. Crown Svo. 18*. MISCELLANEOUS WORKS & POPULAR METAPHYSICS. Arnold's (Dr. Thomas) Miscellaneous Works. Svo. 7*. Gd. Bain's Emotions and the Will. Svo. 15*. — Mental and Moral Science. Crown Svo. 10*. Gd. Or separately : Part I. Mental Science, 6*. Gd. Part II. Moral Science, 4*. Gd. — Senses and the Intellect. Svo. 15*. Buckle's Miscellaneous and Posthumous Works. 3 vols. Svo. 52*. Gd. London, LONGMANS & CO. General Lists of New Works. Carpenter on Mesmerism, Spiritualism, &c. Crown 8vo. 5s, Conington's Miscellaneous Writings. 2 vols. 8vo. 28*. Fronde's Short Studies on Great Subjects. 3 vols, crown 8vo. 18s. Qennan Home Life ; reprinted from Eraser's Magazine. Crown 8vo. 6*. Hume's Essays, edited by Greene & Grose. 2 vols. 8vo. 285. — Treatise of Human Nature, edited by Green & Grose. 2 vols. 8vo. 28«. Kirkman's Philosophy "Without Assumptions. 8vo. 10s. 6d. Macaulay's Miscellaneous Writings. 2 vols. 8vo. 21«. 1 vol. crown 8vo. 4*. 6c?. — Writings and Speeches. Crown 8vo. 6s. Mill's Analysis of the Phenomena of the Hmnan Mind. 2 vols. 8vo. 28*. — Subjection of Women. Crown 8vo. 6*. MUller's (Max) Chips from a German Workshop. 4 vols. 8vo. 58*. Mullinger's Schools of Charles the Great. 8vo. 7s. 6d. Owen's Evenings with the Skeptics. Crown 8vo. [Just ready. Rogers's Defence of the Eclipse of Faith Fcp. 8vo. 3s. 6d. — Eclipse of Faith. Fcp. 8vo. 5s. Selections from the Writings of Lord Macaulay. Crown 8vo. 6*. Sydney Smith's Miscellaneous Works. Crown 8vo. 6s. The Essays and Contributions of A. K. H. B. Crown 8vo. Autumn Holidays of a Country Parson. 3*. 6d. Changed Aspects of Unchanged Truths. 3*. 6d. Common-place Philosopher in Town and Country, o*. 6d. Counsel and Comfort spoken from a City Pulpit. 3*. 6d. Critical Essays of a Country Parson. 3*. Gd. Graver Thoughts of a Country Pardon. Three Series, 3*. 6d. each. Landscapes, Churches, and Moralities. 3«, 6d. Leisure Hours in Town. 3*. 6d. Lessons of Middle Age. 3^. 6d. Present-day Thoughts. 3*. Gd. Eecreations of a Country Parson. Two Series, 3.5. Gd. each. Seaside Musings on Sundays and Week-Days, '3s. Gd. Sunday Afternoons in the Parish Church of a University City. 3*. Gd. Wit and Wisdom of the Eev. Sydney Smith. 16mo. 3*. Gd. ASTRONOMY, METEOROLOGY, POPULAR GEOGRAPHY &c. Dove's Law of Storms, translated by Scott. 8vo. 10*. 6d. Hartley's Air and its Relations to Life. Small 8vo. 6s. Herschel's Outlines of Astronomy. Square crown 8vo. 12*. Keith Johnston's Dictionary of Geography, or Gazetter. Svo. 42*. Neison's Work on the Moon. Medium 8vo. 31*. Gd. Proctor's Essays on Astronomy. Svo. 12*. — Larger Star Atlas. Folio, 15*. or Maps only, 12*. Gd. — Moon. Crown 8vo. 15*. — New Star Atlas. Crown Svo. 5*. — Orbs Around Us. Crown Svo. 7*. 6d. — Other Worlds than Ours. Crown Svo. 10*. Gd. — Saturn and its System. Svo. 14*. — Sun. Crown Svo. 14*. — Transits of Venus, Past and Coming. Crown Svo. 8*. 6d. — Treatise on the Cycloid and Cycloidal Curves. Crown Svo. 10*. 6(/. London, LONGMANS & CO. General Lists of New Works. Proctor's Universe of Stars. 8vo. 10*. 6d. Schellen's Spectrum Analysis. 8vo. 28*. Smith's Air and Rain. 8vo. 24*. The Public Schools Atlas of Ancient G-eography. Imiwrial Svo. 7s. 6d. — — — Atlas of Modern Geography. Imperial Svo. 5*. "Webb's Celestial Objects for Common Telescopes. Crown Svo. 7*. 6d. NATURAL HISTORY 8c POPULAR SCIENCE. Ai-nott's Elements of Physics or Natural Philosophy. Crown Svo. 12s. 6d. Brando's Dictionary of Science, Literature, and Art. 3 vols, medium Svo. 63*. Decaisne and Le Maout's Greneral Sj'stem of Botany. Imperial Svo. 31*. Gd. Evans's Ancient Stone Implements of Great Britain. Svo. 28*. Ganot's Elementary Treatise on Physics, by Atkinson. Large crown Svo. 15*. — Natiu-al Philosophy, by Atkinson. Crown Svo. 7*. 6d. Grove's Coirelation of Physical Forces. Svo. 15*. Hartwig's Aerial "World. Svo. 10*. Gd. — Polar World. Svo. 10*. Gd. — Sea and its Living Wonders. Svo. 10*. Gd. — Subterranean World. Svo. 10*. Gd. — Tropical World. Svo. 10*. 6d. Haughton's Principles of Animal Mechanics. Svo. 21*. Heer's Primaeval World of Switzerland. 2 vols. Svo. 28*. Helmholtz's Lectures on Scientific Subjects. 8vo. 12*. Gd. Hehnholtz on the Sensations of Tone, by EUis. Svo. 36*. Hemsley's Handbook of Trees, Shrubs, & Herbaceous Plants. Medium Svo. 12*. Hullah's Lectures on the History. of Modern Music. Svo. 8*. Gd. — Transition Period of Musical History. Svo. 10*. Gd. Keller's Lake Dwellings of Switzerland, by Lee. 2 vols, royal Svo. 42*. Kirby and Spence's Introduction to Entomology. Crown Svo. 5*. Lloyd's Treatise on Magnetism. Svo, 10*. Gd. — — on the Wave-Theory of Light. Svo. 10s. Gd. Loudon's Encyclopsedia of Plants. Svo. 42*. Lubbock on the Origin of Civilisation & Primitive Condition of Man. Svo. 18*. Nicols' Puzzle of Life. Crown Svo. 3*. Gd. Owen's Comparative Anatomy and Physiology of the "Vertebrate Animals. 3 vols. Svo. 73*. Gd. Proctor's Light Science for Leisure Hours. 2 vols, crown Svo. 7*. Gd. each. Rivers's Rose Amateur's Guide. Fcp. Svo. 4*. Gd. Stanley's Familiar History of Birds. Fcj). Svo. 3*. Gd. Text-Books of Science, Mechanical and Physical. Abney's Photography, small Svo. 3*. Gd. Anderson's Streng-th of Materials, 3*. Gd. Armstrong's Organic Chemistry, 3*. Gd. Barry's Railway Appliances, 3*. Gd. Bloxam's Metals, 3*. Gd. Goodeve's Elements of Mechanism, 3*. Gd. — Principles of Mechanics, 3*. Gd. Gore's Electro-Metallurgy, 6*. Griffin's Algebra and Trigonometry, 3*. Gd. London, LONGMANS & CO. General Lists of New Works. Text-Books of Science— continued. Jeukiu's Electricity and Magnetism, 3^. 6d. Maxwell's Theory of Heat, 3s. 6d. MerrifieJd's Technical Arithmetic and Mensuration, 35. 6d. Miller's Inorganic Chemistry, 3*. Gd. Pi-eece & Sivewright's Telegrapliy, 3s. 6d. Shelley's Workshop Appliances, 3s. 6d. Thom6'3 Structural and Physiological Botany, 6*. Thorpe's Quantitative Chemical Analysis, 4*. M. Thorpe & Muir's Qualitative Analysis, 3*. 6d. Tilden's Chemical Philosophy, 3*. 6d. Unwin's Machine Design, 3^. Gd. "Watson's Plane and Solid Geometry, 3^. 6i. Tyndall on Sound. Crown 8vo. 105. Gd. — Contributions to Molecular Physics. 8vo. 16*. — Fragments of Science. Crown 8vo. 10*. 6c?. — Lectures on Electrical Phenomena. Crown Svo. Is. sewed, 1*. Gd. cloth. — Lectures on Light. Crown Svo. 1*. sewed. Is. Gd. cloth, — Lectures on Light delivered in America. Crown Svo. 7s. Gd. — Lessons in Electricity. Crown Svo. 25. Gd. Woodward's Geology of England and Wales. Crown Svo. 14*. Wood's Bible Animals. With 112 Vignettes. Svo. 145. — Homes Without Hands. Svo. 14*. — Insects Abroad. Svo, 145. — Insects at Home. With 700 Illustrations. Svo. 145. — Out of Doors, or Articles on Natural History. Crown Svo. 75. Gd. — Strange Dwellings. With 60 Woodcuts, Crown Svo. 75. Gd. CHEMISTRY 8c PHYSIOLOGY. Auerbach's Anthracen, translated by ^Y. Crookes, F.R.S. Svo. 12.$. Buckton's Health in the House ; Lectures on Elementary Physiology. Fcp. Svo. 2s. Crookes's Handbook of Dyeing and Calico Printing. Svo. 425. — Select Methods in Chemical Analysis. Crown Svo. 125. Gd. Kingzett's Animal Chemistry. Svo. [/« Ihe press. — History, Products and Processes of the Alkali Trade. Svo. 125. Miller's Elements of Chemistry, Theoretical and Practical. 3 vols. Svo. Part I. Chemical Physics, 165. Part II. Inorganic Chemistry, 2l5. Part III. Organic Chemistry, New Edition in the press. Watts's Dictionai-y of Chemistry. 7 vols, medium Svo. £10. 165. Gd. THE FINE ARTS & ILLUSTRATED EDITIONS. Doyle's Fairyland ; Pictures from tlse Elf -World, Folio, 155. Jameson's Sacred and Legendary Art. 6 vols, square crown Svo. Legends of the Madonna. 1 vol. 2l5. — — — Monastic Orders. 1 vol. 2l5. — — — Saints and Martyrs. 2 vols. 3l5. Gd. — — — Saviour. Completed by Lady Eastlake. 2 vols. 425. London, LONGMANS & CO. General Lists of New Works. Longman's Three Cathedrals Dedicated to St. Paul. Square crown 8vo. 21». Macaulay's Lays of Ancient Rome. "With 90 Illustrations. Pep. 4to. 21*. Macfarren's Lectures on Harmony. 8vo. 12s. Miniature Edition of ilacaulay's Lays of Ancient Rome. Imp. 16 mo. 10«. 6d. Moore's Irish Melodies, With 161 Plates by D. Maclise, R.A. Super-royal 8vo. 2l5. — Lalla Rookh. Tenniel's Edition. With 68 Illustrations. Pep. 4to. 21*. Redgrave's Dictionary of Artists of the English School. 8vo. 16*. THE USEFUL ARTS, MANUFACTURES &c. Bourne's Catechism of the Steam Engine. Pep. 8vo. 6*. — Handbook of the Steam Engine. Pep. 8vo. 9.?. — Recent Improvements in the Steam Engine. Pep. 8vo. 6*. — Treatise on the Steam Engine. 4to. 42*. Cresy's Encyclopaedia of Civil Engineering. 8vo. 42«. Cnlley'a Handbook of Practical Telegraphy. 8vo. 16*. Eastlake's Household Taste in Pumiture, &c. Square crown 8vo. lis. Pairbairn's Useful Information for Engineers. 3 vols, crown 8vo. 31«. Qd. — Applications of Cast and Wrought Iron. 8vo. 16«. G-wUt's EncyclopaBdia of Architecture. Svo. 525. 6d. Hobson's Amateur Mechanics Practical Handlwok. Crown 8vo. 2*. dd. Hoskold's Engineer's Valuing Assistant. 8vo. 31s. 6d. Kerl's Metallurgy, adapted by Crookes and Rbhrig. 3 vols. 8vo. £4. Ids. Loudon's Encyclopaedia of Agriculture. 8vo. 21*. — — — Gardening. Svo. 21*. Mitchell's Manual of Practical Assaying. 8vo. 31*. 6d. Northcott's Lathes and Turning. Svo. IS*. Payen's Industrial Chemistry, translated from Stohmann and Engler's German Edition, by Dr. J. D. Barry. Edited by B. H. Paul, Ph.D. Svo. 42?. Stoney's Theory of Strains in Girders. Roy. Svo. 36*. Ure's Dictionary of Arts, ZManufactures, & Mines. 3 vols, medium Svo. £5. 5*. Supplementary Volume of Recent Improvements. 42*. (Nearly ready.) RELIGIOUS & MORAL WORKS. Arnold's (Rev. Dr. Thomas) Sermons. 6 vols, crown Svo. 5*. each. Bishop Jeremy Taylor's Entire Works. With Life by Bishop Heber. Edited by the Rev. C. P. Eden. 10 vols. Svo. £5. 5*. Boultbee's Commentary on the 39 Articles. Crown Svo. 6*. Browne's (Bishop) Exposition of the 39 Articles. Svo. 16*. Colenso on the Pentateuch and Book of Joshua. Crown Svo. 6*. Colenso's Lectures on the Pentateuch and the Moabite Stone. Svo. 12*. Conybeare & Howson'sLife and Letters of St. Pau.1 : — Library Edition, with all the Original Illustrations, Maps, Landscapes on Steel, Woodcuts, &c. 2 vols. 4to. 42*. Intermediate Edition, \vith a Selection of Maps, Plates, and Woodcuts. 2 vols, square crown Svo. 21*. Student's Edition, revised and condensed, with 46 Illustrations and Maps. 1 vol. crown Svo. 9*. London, LONGMANS & CO. General lists of New Works. D'Aubigng's Eeformation in Europe in the Time of Calvin. 8 vols. 8vo. £6. 125. Drummond's Jewish Messiah. 8vo. 15*. Ellicott's (Bishop) Commeutary on St, Paul's Epistles. Svo. Galatians, 85. 6(7. Ephesians, 8*. 6d. Pastoral Epistles, 10*. Gd. Philippians, Colossians, and Philemon, 10*. Gd. Thessalonians, 7s. Gd. Ellicott's Lectm-es on the Life of our Lord. 8vo. 12*. Ewald's History of Israel, translated by Carpenter. 5 vols. Svo. 63*. — Antiquities of Israel, translated by Solly. 8vo. 12*. Gd. Goldziher's Mythology among the Hebrews. Svo. 16*. Griffith's Behind the Veil ; an Outline of Bible Metaphysics. Svo. 10*. Gd. Jukes's Types of Genesis. Crown Svo. 7*. Gd. — Second Death and the Restitution of all Things. Crown Svo. 3*. Gd. Kalisch's Bible Studies. Part I. the Prophecies of Balaam. Svo. 10*. Gd. Keith's Evidence of the Truth of the Christian Religion derived from the Fulfil- ment of Prophecy. Square Svo. 12*. Gd. Post Svo. 6*. Kuenen on the Prophets and Prophecy in Israel. Svo. 21*. Lyra Germanica. Hymns translated by Miss Winkworth. Ecp. Svo. 5*. Manning's Temporal Mission of the Holy Ghost. Svo. 8*. Gd. Martineau's Endeavours after the Christian Life. Crown Svo. 7*. Gd. — Hymns of Praise and Prayer. Crown Svo. 4*. Gd. 32mo. 1*. Gd. — Sermons ; Hom-s of Thought on Sacred Things. Crown Svo. 7*. 6^/. Mill's Three Essays on Religion, Svo. 10*. Gd. Monsell's Spiritual Songs for Sundays and Holidays. Fcp. Svo. 5*. ISmo. 2*. Miiller's (Max) Lectures on tlie Science of Religion. Crown Svo. 10*. Gd. O'Conor's New Testament Commentaries. Crown Svo. Epistle to the Romans. 3*. 6c;. Epistle to the Hebrews, 4*. Gd. St, John's Gospel, 10*. Gd. Passing Thoughts on Religion. By Miss Sewell. Fcp. Svo. 3*. Gd, Sewell's (Miss) Preparation for the Holy Communion. 32mo. 3*. Shipley's Ritual of the Altar. Imperial Svo. 42*. Supernatm-al Religion. 3 vols. Svo. 38*. Thoughts for the Age. By Miss Sewell. Fcp, Svo, 3*, Gd. Vaughan's Trident, Crescent, and Cross ; the Religious History of India, 8vo,9*.6./, Whately's Lessons on the Christian Evidences. 18mo. Gd. White's Four Gospels in Greek, with Greek-English Lexicon, 32mo, 5*. TRAVELS, VOYAGES &c. Ball's Alpine Guide. 3 vols, post Svo, with Maps and Illustrations :— I. Western Alps, 6*. Gd. II. Central Alps, 7s. Gd. III. Eastern Alps, 10*. Gd. Or in Ten Parts, 2*. Gd. each. Ball on Alpine Travelling, and on the Geology of the Alps, 1*. Each of the Three Volumes of the Alpine Guide may be had with this Introduction prefixed. price 1*. extra. Baker's Rifle and the Hound in Ceylon. Crown Svo. 7*. Gd. — Eight Years in Ceylon. Crown Svo. 7*. Gd. Brassey's Voyage in the Yacht ' Sunbeam.' Svo, 21*, Edwards's (A, B.) Thousand Miles up the Nile. Imperial Svo. 42*. Edwards's (M. B.) Year in Western France. Crown Svo. 10*. Gd. London, LONaMANS & CO, 10 General Lists of New Works. Evans's Through Bosnia and Herzegovina during the Insurrection. 8vo. 18«. — niyrian Letters. 8vo. {Nearly ready.) Grohman's Tyrol and the Tyrolese. Crown 8vo. Gs. HinchlifE's Over the Sea and Far Away. Medium 8vo. 2\s. Indian Alps (The). By a Lady Pioneer. Imperial 8 vo. 425. Lefroy's Discovery and Early Settlement of the Bermuda Islands. Vol. I. Eoyal 8vo. 30*. Packe's Guide to the Pyrenees, for Mountaineers. Crown 8vo. Is. M. The Alpme Club Map of Switzerland. In four slieets. A2s. Wood's Discoveries at Ephesus. Imperial 8vo. 635. WORKS OF FICTION. Becker's Charicles ; Private Life among the Ancient Greeks. Post 8vo. Is. M. — GaUus ; Roman Scenes of the Time of Augustus. Post 8vo. 7*. 6d. Cabinet Edition of Stories and Tales by Miss SeweU :— Amy Herbert, 2s. 6d. Cleve HaU, 2*. 6d. The Earl's Daughter, 2*. 6d. Experience of Life, 2*. 6d. Gertrude, 25. &d. Novels and Tales by the Right Hon. the Earl of Beaconsfleld, complete in Ten Volumes, crown 8vo. price £3. Ivors, 25. 6d. Katharine Ashton, 25. M. Laueton Parsonage, 35. 6c7. Margaret Percival, 35. &d. Ursula, 35. M. Cabinet Edition, Lothair,' 65. Coningsby, 65. Sybil, 65. Tancred, Qs. Venetia, 6s. The Atelier du Lys ; or, an Art Student in the Reign of ' Mademoiselle Mori.' Crown 8vo. 65. Henrietta Temple, 65. Contarini Fleming, 65. Alroy, Ixion, &c. 65, The Young Duke, &c. 65, Vivian Grey, 65. Terror. By the Author The Modem Novelist's Library. Each Work in crown 8vo. A Single Volume, complete in itself, price 25. boards, or 25. Qd. cloth :— By Lord Beaconsfleld. Lothair. Coningsby. Sybil. Tancred. Venetia. Henrietta Temple. Contarini Fleming. Alroy, Ixion, &c. The Young Duke, &c. Vivian Grey. By Anthony TroUope. Barchester Towers. The Warden. By the Author of ' the Rose Garden. Unawares. By Major Whyte-Melville. Digby Grand. General Bounce. Kate Coventry. The Gladiators. Good for Nothing. Holmby House. The Interpreter. The Queen's Maries. By the Author of ' the Atelier du Lys Mademoiselle Mori. By Various Writers. Atherstone Priory. The Burgomaster's Family. Elsa and her Vulture. The Six Sisters of the Valley. Wliispers from Fairy Land. By the Right Hon. E. H. KnatchbuU-Hugessen M.P. With Nine lUustrations. Crown 8vo. 35. M. Higgledy-piggledy ; or. Stories for Everybody and Everybody's Children. By the Right Hon. E. M. Knatchbull-Hugessen, M.P. With Nine Illustrations from Designs by R. Doyle. Crown 8vo. 35. 6(7. London, LONGMANS & CO. General Lists of New Works. 11 POETRY 8c THE DRAMA. Bailey's Festus, a Poem. Crowu 8vo. 12*. 6(/. Bowdler's Family Shakspeare. Medium 8vo. lis. 6 vols. fcp. 8vo. 21*. Conington's ^iieid of Virgil, translated into English Verse. Crown Svo. 9*. Cayley's Iliad of Homer, Homometrically translated. Svo. 12*. 6d. Ingelow's Poems. First Series. Illustrated Edition. Fcp. 4to. 21*. Macaulay's Lays of Ancient Rome, with Ivry and the Armada. 16mo. 3*. 6d. Poems. By Jean Ingelow. 2 vols. fcp. Svo. 10*. First Series. ' Divided,' ' The Star's Monument,' &c. 5*. Second Series. ' A Story of Doom,' ' Gladys and her Island,' &c. 6*. Southey's Poetical Works. Medium Svo. 14*. Yonge's Horatii Opera, Library Edition. Svo. 21*. RURAL SPORTS, HORSE & CATTLE MANAGEMENT &c. Blaine's Encydopsedia of Rural Sports. Svo. 21*. Dobson on the Ox, his Diseases and their Treatment. Crown Svo, 7s. 6d. Fitzwygram's Horses and Stables. Svo. 10*. Gd. Francis's Book on Angling, or Treatise on Fishing. Post Svo. 15*. Malet's Annals of the Road, and Nimrod's Essays on the Road. Medium Svo. 21*. Miles's Horse's Foot, and How to Keep it Sound. Imperial Svo. 12*. 6d. — Plain Treatise on Horse-Shoeing. Post Svo. 2*. 6d. — Stables and Stable-Fittings. Imperial Svo. 1-5*. — Remarks on Horses' Teeth. Post Svo. 1*. 6^^;. Moreton on Horse- Breaking. Crowu Svo. 5*. NevQe's Horses and Riding. Crown Svo. 6*. Reynardson's Do^vn the Road. Medium Svo. 21*. Ronalds's Fly -Fisher's Entomology. Svo. 14*. Stonehenge's Dog in Health and Disease. Square crown Svo. 7s. 6d. — Greyhound. Square cro^^•n Svo. 15*. Youatt's Work on the Dog. Svo. 12*. 6d. — — — — Horse. Svo. 6*. "Wilcocks's Sea-Fisherman. Post Svo. 12*. Gd. WORKS OF UTILITY & GENERAL INFORMATION. Acton's Modern Cookery for Private Families. Fcp. Svo. 6s. Black's Practical Treatise on Brewing. Svo. 1 0*. Gd. Bull on the Maternal Management of Children. Fcp. Svo. 2*. Gd. Bull's Hints to Mothers on the Management of their Health during the Pregnancy and in the Lying-in Room. Fcp. Svo. 2*. Gd. Campbell-Walker's Correct Card, or How to Play at Whist. 32mo. 2*. Gd. Crump's English Manual of Banking. Svo. 15*. Longman's Chess Openings. Fcp. Svo. 2*. Gd. Macleod's Theory and Practice of Banking. 2 vols. Svo. 26*. — Elements of Banking. Crown Svo. 7*. Gd. London, LONGMANS & CO. M'Culloch's Dictionary of Commerce and Commercial Navigation. 8vo. 63*. Maunder's Biographical Treasury. Fcp. 8vo. 65. — Historical Treasury. Fcp. 8vo. 6*. — Scientific and Literary Treasury. Fcp. 8vo. 6,?. — Treasury of Bible Knowledge. Edited by the Rev. J. Avre, M. A. Fcp. 8vo. 6*. — Treasury of Botany. Edited by J. Lindley, F.R.S. and T. Moore, F.L.S. Two Pai'ts, fcp. 8vo. 12*. — Treasury of Greography. Fcp. 8vo. 6s. — Treasury of Knowledge and Library of Reference. Fcp. 8vo. 6*. — Treasury of Natural History. Fcp. 8vo. 6s. Pewtner's Compreliensive Si)ecifier ; Building- Artificers' "Work. Conditions and Agreements. Crown 8vo. 6s. Pierce's Three Hundred Chess Problems and Studies. Fcp. 8vo. 7*. Gd. Pole's Theory of the Modern Scientific G-ame of Whist. Fcp. 8vo. 2s. 6d. The Cabinet La^^'J'e^ ; a Popular Digest of the Laws of England. Fcp. 8vo. 9s. WUlich's Popular Tables for ascertaining the Value of Property. Post 8vo. 10s. WUson's Resources of Modem Countries 2 vols. 8vo. 24*. MUSICAL WORKS BY JOHN HULLAH, LLD. Chromatic Scale, with the Inflected Syllables, on Large Sheet. Is. 6d. Card of Chromatic Scale. Id. Exercises for the Cultivation of the Voice. For Soprano or Tenor, 2s. 6d. Grammar of Musical Harmony. Royal 8vo. 2 Parts, each 1*. 6i. Exercises to Grammar of Musical Harmony. Is. Grammar of Counterpoint. Part I. super-royal 8vo. 2s. 6d. Hullah's Manual of Singing. Parts I. & II. 2s. 6d. ; or together, 5*. Exercises and Figures contained in Parts I. and II. of the Manual. Books I. & II. each Sd. Large Sheets, containing the Figures in Part I. of the Manual. Nos. 1 to 8 in a Parcel. 6.?. Large Sheets, containing the Exercises in Part I. of the Manual. Nos. 9 to 40, in Four Parcels of Eight Nos. each, per Parcel. 6*. Large Sheets, the Figures in Part II. Nos. 41 to 52 in a Parcel, 95. Hymns for the Young, set to Music. Royal 8vo. 8d. Infant School Songs. 6d. Notation, the Musical Alphabet. Crown 8vo. 6d. Old English Songs for Schools, Harmonised. 6d. Rudiments of Musical Grammar. Royal 8vo. 3*. School Songs for 2 and 3 Voices. 2 Books, 8vo. each 6d. Time and Tune in the Elementary School. Crown 8vo. 2*. 6d. Exercises and Figures in the same. Crown 8vo. 1*. or 2 Parts, 6d each. London, LONGMANS & CO. Spottiswoode d: Co., Printers, New-street Sqtinre, London^ 14 DAY USE RETURN TO Cff'AR^'^DrF^T OWED This book is due on the last date stamped below, on the date to which renewed. Renewed bo6ks are subject to immediate recall. ^^ f^ ^ IN S T ACKS I- LB 2 61957 RuTC'D LD J.'.:: 1 5 1962 "^^ <^^^ REC'D LD Z^'itK^^^^Trt t-f^ n& i^ tPiJt^ ouom - HJan'SgJWW m2 7 1970 56 m^' Jim d^^ %^-^ LD-21-100m-6,'56 (B9311sl0)476 General Library University of California Berkeley YB U.C.BERKELEY LIBRARIES CD4afiS3STS