ftl !;hi«iHiiiuuiuiiiMiiUiHiu Cornell University Library The original of tliis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924031787561 Cornell University Library arW37667 A treatise on versification / 3 1924 031 787 561 oiin,anx TREATISE VERSIFICATION. EEV. E"W. EVANS, TICAR OF HEVEBSHAM, AND LATE FELLOW OF TRINITY COLLEGE, CAMBRIDGE. LONDON: FRANCIS & JOHN RIVINGTON, ST. Paul's church yard, and Waterloo place. 1852. -S-Ws—E-iS-i /^ y~\ i~\ ;'•";• fi. ,! ?■;' a j \ s f^'- %^K.. ^^ v.„ . '< lis; LONDON : filLBERT & EIVINGTON, PRINTEES, ST. John's square. PKEFACE. The following Treatise has long lain by the Author, in hopes that its subject might be taken up by some other hand, more able to discuss the general principles which should give a common basis to ancient and modem Versification, and thus introduce some breadth of view and interest into a subject which seems to have been con- sidered with partial glimpses, and made more dry and duU than any other. The confusion hence arising is exceedingly great, as may be seen in the attempts to introduce the ancient metres into our modern tongues, which have no prosody of quantity. a2 PREFACE. No such work having appeared, as far as the Author's hnowledge extends, he has been induced to put the Treatise in order for publication. He is moved also by the consideration of the very essential use which a knowledge of the prin- ciples of versification, involving as they do the elements of recitation, is tb correct and good reading ; both as to the clear and firm enuncia- tion of the syllables, and the management of the pauses. Without these no flexibility can be ob- tained, but all will be an indistinct monotonous drawl, varied perhaps by unmeaning starts. It has been remarked that poets always write' good prose. The reason is obvious. And not less obvious is the reason why good readers of poetry are always good readers of prose : nor can there be a good reader of prose, fully equal to every occasion, who has not first made himself a good reader of poetry. His recitation will not have been surely founded on the principles of the language, and must, therefore, be uncertain at best. ■ Unfor- tunately good reading is quite as scarce as every PREFACE. V other good thing. And, therefore, it seems no unprofitable work to promote it, especially when it is considered on what serious occasions it comes into exercise, and what a grievous hin- drance bad reading then is. CONTENTS. CHAPTEE I. Fasz On the Nature of Verse in General ... .1 CHAPTEE II. On Quantity 12 CHAPTEE III. On Stress 19 CHAPTEE IV. On tlie Construction of a Verse 27 CHAPTEE V. On the Pause 40 CHAPTEE VI. On the Pauses of the Hexameter > .... 52 CHAPTEE Vn. On the Pauses of the Trimeter 64 Vm CONTENTS. CHAPTEE VIII. Page On the Pauses of the Alexandrine 76 CHAPTEE IX. On the Pauses of the Heroic 82 CHAPTEE X. On the Pauses of the various Tetrameters .... 91 CHAPTEE XI. On the Expression due to the Articulation of Syllables . 97 CHAPTEE XII. On Alliteration 107 CHAPTEE XIII. On Rhyme 113 CHAPTEE XIV. On the Modem Recitation of Ancient Verse . . . 123 CHAPTEE XV. On the Accentuated Hexameter 131 CHAPTEE XVI. On the Modem Hexameter 138 CHAPTEE XVII. ■ On Stanzaic Poetry 146 CHAPTEE XVIII. On Lyric Poetry 162 TREATISE ON VERSIFICATION. CHAPTER I. ON THE NATURE OP VERSE IN GENERAL. 1. The word Verse has been derived from the Latin tongue, in which it means in general a turning, and, in particular, the turning of the ox and plough at the end of the furrow. Thence, by an obvious figure, it was applied to a written line, whether in prose or verse; for in the primitive fashion of writing, the letters of the consecutive line turned in a direction opposite to that of the preceding, as does the horse or ox in the consecu- tive furrow. But apart from this particular cus- tom, a stone or parchment crossed with lines could not but put the farmers, as all then were, very forcibly in mind of a field crossed with furrows, B 2 ON THE NATURE and lead them to the figurative language. Thus we are referred, by this meaning of the word, not to any qudity of the language employed, but sim- ply to the quantity of space occupied on the leaf, stone, or tablet. And since poetry was the main subject of transcription in the interval between the periods of sculptured monuments and familiar prose writing, the term Verse came to be appro- priated to its measured line, and in this sense only has entered ours and every other modern language. 2. Here then we find the first article of its de- finition, which is, that it is of an assigned extent. But further, this assigned extent, on being re- peated, leads us to demand a similarity of struc- ture ; and the marks of this similarity must of course be expressed by the recitation, through which alone, originally, verse was conveyed to the mind, and to which, at all times, even while read only by the eye, it is referred by the mind, and so regarded by it exclusively according to the effects on the ear. But these are constant and essential only as to the individual words : they are variable as to sentences. Now the recitation of words involves but three things : (1) A certain length of time given to the pronunciation of each syllable. (2) The pronunciation of each syllable in a certain key. (8) The pronunciation of it in a certain loud- ness of tone. The similarity of structure, there- OF TERSE IN GENERAL. 3 fore, must be found in the repetition of some one of these, and to these three elements, therefore, and no where else, we must look for the constant basis of versification, which, accordingly, may be defined as — A series of words, which is of definite extent, and follows an arrangement which depends on the regulated recurrence of a syllable which is pecu- liarly marked according to some one of the fore- mentioned elements. The extent is determined by the number of in- tervals and returns thus formed, for example in the verses, — Murmviring, sighing, and sorrowing, To tlie woods, to the mountains, and streams. The syllables follow according to the arrangement of the return of the stress at the interval of every two ; and the verses are equal in length, each con- sisting of three returns and three intervals, only in a reverse order in either line Thus verse is entirely independent of the in- ward and mental part of language, so that even utter nonsense may be conveyed in the most har- monious verse ; and, depending on the invariable properties of the voice alone, it has nothing really to do with alliteration or rhyme, or any of the variable accidents of mere articulation, though b2 4 ■ ON THE NATURE these may be taken additionally into the account for the sake of ornament. And it is from the numher of the returns, and from the measure of the intervals, and from the proportions thus pre- sented, that the terms numbers, measv/re, rhythm, time, harmony, and the like, here become ap- plicable to verse. We will now proceed to discuss these elements of versification. 3. The first is the time taken up in the pronun- ciation of a syllable. Thus, if in any language its syllables were divisible into two or more classes, according to two or more invariable times of pro- nouncing them, then a return might be made by the syllables of one time recurring at the interval of syllables of another time, or even of two other times, as thus, AaAaAa, &c., or even AaaAaaAaa, &c. The latter, however, of these schemes is too artificial for any known language. It would not indeed be too complex for the eye, which can com- prehend a whole in one glance, and dwell long enough upon it for comparison of parts, as in the survey of specimens of embroidery and archi- tecture ; but it would scarcely be appreciated by the ear, which can bring but a portion present be- fore the mind at a time, the preceding part exist- ing but in memory, and the succeeding anticipated in thought from it. The only languages, there- fore, that have employed this basis for verse, which OF VERSE IN OENERAL. 5 are the Greek and Latin, have been content with acknowledging hut two times, and those in the simplest proportion of one to two, so that in the word musa, the pronunciation of the first syllable shall take up twice as much time as does that of the second. Our word muses exemplifies a similar difierence of times. The former of these is called the long time, the latter the short time. The return is naturally denoted by the former, as by the more important, the interval by the latter ; and a verse will be formed thus : — Musa quK meas ad aures, Muses all divine and holy, where the return is double of the interval. Or thus, Experientia parcis, where the return and interval are equal. Such a basis has proved quite sufficient in the supply of variety for every possible use of verse, and indeed far more sufiicient than the two put together that remain for mention. Hence the word quantity has been applied to syllables, as denoting the quantity of time which their pronunciation occu- pies. 4. The second is the key of voice in which a syllable is pronounced. Thus, if any language had b3 6 ON THE NATURE its syllables divisible into two or more classes, ac- cording to the respective keys, higher or lower, in which by invariable usage they were pro- nounced, here again were a basis for versification. Suppose a certain number were pronounced in the note F, another in A, another in 0, then syllables would admit of the arrangement, FACFAC, &c., together with the varieties arising from the permu- tations of the combination FAC. And there would be just the same resources for return and interval as before ; but here again, also, even the most musical of languages, the Greek and Latin, do not recognize more than two classes ; the one, of syl- lables pronounced in a lower tone, called the grave, the other, of syllables pronounced in a higher tone, called the acute ; the latter but on one syl- lable of the word, the former occupying all the remainder, as in calamitas ; and the harmony re- sulting, and the flexibility of the instrument are very inferior to what is experienced in the former case ; for although no return can be better marked than that which is afibrded by the syllable on which the acute faUs, yet the means of supplying the interval are very imperfect, since this must now be regulated by the number of syllables on which the grave falls, there being no other measure of them, and not by the intervening time, which may occupy the place of one or more syllables. OF VERSE IN GENERAL. 7 For example, in amnibus and amnes the interval is the same in time but different in number, and in ^mnes and ^mnis different in time but the same in number. Hence this basis is much inferior to the former, inasmuch as, through its limitation of the number of syllables in the interval, it excludes all that expansion and contraction which gives such expression according to the former basis, allowing the hexameter to roU on through seven- teen syllables, or stop at thirteen. Besides, the interval is subjected to a foreign influence; namely, the number, as well as the gravity, of the syllables, and thus the simplicity and symmetiy of the for- mer basis does not exist. But there is also an insuperable difficulty attend- ing the construction of verse upon it, and which is the greater in proportion to the polysyllabic genius of the language. According to the former basis, we may have more than one return in a word, as in candidos, and more than one interval, as in silebit ; and thus we have little difficulty in finding words to foUow each other according to any required metrical arrangement. But here one syllable only in a word can make the return, as in animadversi6nes, while an unmanageable number may form the interval on one or both sides of it, as in this very word, which cannot enter any or- dinary system according to the strict definition of B 4 8 ON THE NATUEE verse before given, if the interval be always the same ; and that, except in the lyrical measures, it must be. In a language, therefore, which abounds in such words as calamitatibus, meditatidnes, how can we form the most common system of verse, that of alternate syllables of return and interval like our heroic measure ? In short, the thing were impossible (as a whole), even in a language as abundant in short words as our own. Let any one try to make several trochaic, iambic, or even dactylic lines (according to accent) on this basis in Latin or Greek, and he wiU quickly find himself at a loss. For example, lUe maguos per terrdres. Simplest as this metre is of all, yet it excludes all words of more than three syllables, and even of these aU that have not the accent on the mid- dle. Such words as t^rritos, ayavdHv, which im- pose an interval of two syllables, could not enter. Such a line as, Inaccessas regiones, depends on quite another principle, to which we now come. 5. The third, which is the stress laid upon a syllable. If there were in a language two classes of syllables, according to one of which they were pronounced in a louder tone, and according to the OF VERSE m GENERAL. 9 other in a softer, here would still be a resource for return and interval, but with much less music than the last affords, and with the same insuper- able difficulty, if a, word admit the stress but on one syllable, which is laid down for the rule in every language. Yet this is the only basis of ver- sification in every modern language. How it has been pressed into service we shall see presently. 6. We may observe in passing that the very im- perfect base of similarity of sound has been applied to versification by some barbarous languages, as the Welsh and the Anglo-Saxon ; and therefore, as well as on account of its incidental aid to the true bases, we shall have to take it into consideration. 7. Such therefore are the only legitimate bases available for versification. Not that, however, it would be right to pass over altogether without notice another system also, which some have considered as a species of versification, though excluded by the adoption of the above definition, being based upon the mental part of the lan- guage, not on its pronunciation. It is that com- position of language in which the return is made by a similarity of the syntaxical construction of a sentence, which may run in the parallel or an- tithetical form, giving also the repetition of the idea according to similarity or contrariety. It has been termed rhythmical or measured prose, b5 10 ON THE NATURE and was the only vehicle of Hetrew poetry, if so, we may properly call it. For the sake of illustra- tion, let us refer to the twenty-fourth Psalm, the whole of which runs in the following manner : The earth is the Lord's, and the fulness thereof: The compass of the world, and the dwellers therein. For he hath founded it upon the seas, And prepared it upon the floods, &c. 8. Thus, we are furnished with the exact limits of verse and prose. However poetical in thought, and measured in sentences the latter may he, it cannot on that account make the slightest ap- proach towards verse, unconstructed, as it is, on any regular return of the elements of speech. In fact, whenever such return becomes discernible, we are offended as with an impertinence, and the most musical sentence in prose will ever be the last to remind us of verse. We may, indeed, sometimes find an hexameter, more often a tri- meter, in Greek or Latin prose ; but as being evidently such by mere accident, and according to mere scanning of syllables. For the marked pauses of verse are so wanting, that it requires some curiosity to find out that it has the feet of a complete verse. How many, if not told, would be aware that Tacitus began his Annals with an hexameter : — Urbem Romam a principio Reges habuere. OF VERSE IN GENEEAL. 11 And SO again, however prosaic in thought and expression, a verse may still maintain its dis- tinctive character, and be discerned from the prose, in which quotation may have imbedded it. We will now proceed to examine these three bases in detail Bb" 12 CHAPTER II. ON QUANTITY. 9. The clear pronunciation of every language of ordinary cultivation, requires each syllable to be uttered in a fixed time. That time may be the same for aU syllables, as it is almost entirely, in the Hebrew, pronounced according to the Masoretic rule, and as is too much the case, for poetical capability, at least, in the French. But com- monly it is longer for some, shorter for others. And so constant is this regulation, that any de- viation from it is felt as a vicious barbarism. An Englishman, for example, is offended at hearing his language from the mouth of a Frenchman or Welshman, who seem to pronounce all syllables in the same long time, as in waterr. The proportion of long to short, though reckoned in Greek and Latin in the definite ratio of two to one, cannot, however, be constant, on account of the difference made by the presence of consonants in the syl- lables. Thus the word Christ is sensibly longer ON QUANTITY. 13 than the word cry ; and we find a progression of length in the words gray, grace, grac'd, and in graze, graz'd. The same thing was observed in the Greek, by Dionysius of Halicamassus, who exemplified it in the words 68oe, poBov, Tpoirog, arpoipog. Still, however, the two classes are so distinct, that the number of words which may be ascribed indifier- ently to either, and are therefore termed common, is exceedingly small in any language. And, in- deed, its existence at all is more often owing to poets taking advantage of an older pronunciation, than to any uncertainty about the quantity in their day. And therefore, in modern languages, since their versification depends not upon quan- tity, so that the revival of such old forms is uncalled for, and all left to the practice of the day, there is no such number, except in some few words, which having changed their accent, have happened thereby to change their quantity also. Such with us are words derived from the French, as barrier, wliich is used by Pope with the accent on the last ; and therefore with the last long instead of short ; and one or two more, in which the pronunciation of Chaucer's days may be allowed by poetical licence to linger still. 10. The Greek and Latin had a steady yet flexible pronunciation under their two set times ; and thus, they maintained a golden mean be- 14 ON QUANTITY. tween those languages whicli for the most part employ but one time, and those which, like ours, have times of such various shortness, as even to have no assignable proportion to either a long syllable, or to one another; for what possible proportion can we discern between the times of pronouncing the first syllable of nature, and any of the three last of admirable. This defect is the consequence of the stress, and must affect all languages that employ it, as do all modern. For the nature of the stress is to lengthen the syllable, which is already long, as in nature, and to shorten the syllables which are already short as in ad- mirable^; and thus, a wide gap is created between the longest and the shortest time, and not only all reduction to two times becomes impossible, but the several short times are distinguished more by the practice of a good ear than by any rules that can be found to regulate them. So much has this uncertainty been felt, that some of our English grammarians have asserted, that in nineteen out of twenty of our words it is utterly indifferent whether we pronounce them long or short ; and others have maintained that, ' It should be borne in mind that in our prosody, two conso- nants following a short vowel do not make the syllable long, because we slur the two into one. We do not say ad-mirable, distinctly pronouncing each consonant, and so gaining the time, as do the Welsh and French. ON QUANTITY. 15 at least, the quantity of all our monosyllables is unsettled. Their mistake has arisen from view- ing our language through the medium of the Greek and Latin ; and, therefore, limiting time to hut one proportion ; and, still further, from the confusion which they make between stress and quantity. As it is so necessary to be clear upon this matter, before proceeding further, here fol- lows a table, exhibiting at once the long and short vowels in our language, and according as the latter stand, when not made rapid by falling in the midst of unemphatic syllables, as are the i and a in admirable. 11. LONa. 1. ^ in paU, drawl. 2. A. in. father. 3. A in Matthew, grass. 5. a in late ai in baite ea in great ei in heir. SHORT. in poU, doll. A in mat, gas. f e in her i in stir in done, world 00 in blood u in fun. e in let, bet, regret, in- heritor ea in bread. 16 ON QUANTITY. LONG. 6. e in Bede ee in bleed ea in read i in machine. 7. in pore ou in pour 00. in roar 00 in floor in droll. 8. 00 in rood, woo'd u in rude ue in blue ew in stew. 9. Diphthongs, being aU syllables that have w or y, followed by a simple vowel, as would, yard. And all syllables con- taining u, which is made up of Nos. 6 and 8, and also con- taining % which is made up of Nos. 4 and 6, and all syl- lables containing ou SHOET. i in bid, rid, chin. 00 in wood M in full ou would. ON QtJANTITT. 17 LONG. or ow, which is made up of Nos. 4 and 8, and containing oi, which is made up of Nos. 4 and 6. ew is the only diph- thong expressed ade- quately in letter, be- ing the same with u. 10. Triphthongs, consist- ing of w followed by a diphthong, as in wind, wound (from wind), way. 12. Thus quantity is very distinctly marked in our language, and most especially that of our monosyllables, since there the stress least disturbs the time. And, what may seem strange, the number of long sounds exceeds that of short in the proportion of 7 : 6, putting out even the diphthongs and triphthongs. But then the recur- rence of them is in a very different proportion, that of long being to that of short in the ratio of only 10: 23, owing to so many being unemphatic (77). Thus it would be impossible to construct metre on the basis of quantity in our language. 18 ON QUANTITY. The very simplest measure, that of alternate long and short, would be impracticable to any sufficient extent. For such a purpose the recurrences of long and short should be nearly equal, as we shall see presently to be the case in the Greek and Latin. But we need not repine at a lot which necessarily befalls every modern language, regu- lated and tuned as they are by stress : and the Italian, having much longer words than ours, must have a much greater deluge of unemphatic, and therefore short syllables of ill-proportioned time. Nothing can show more clearly the confusion which has prevailed between quantity and stress than the attempts which have been made to intro- duce the ancient metres into the modern tongues. So prejudiced are we with the stress, that we are in fact utterly indisposed to any attention to long and short, beyond such as is required for correct- ness of pronunciation, and are thus rendered quite insensible to that beauty which is the effect of metrical quantity ; and so entirely do we read the ancient poetry according to stress, that we could not, without such long and painful practice as none has undertaken, recite a single verse according to quantity : and assuredly if we did, then we should astonish the ears of our listeners much more than gratify them. 19 CHAPTER III. ON STRESS. 13. Accent, properly so called, does not exist as a recognized regulation of speech in any modem language ; however some, as the French and Welsh, possess it to a certain degree. But so much is the contrary the case in ours, that the variation between a higher and lower key in speaking is the mark of a rustic and provincial pronunciation. Since then the Greek and Latin made it no basis of versification, we have no need of noticing it any further, after having stated that in those languages, as the ear grew more unrefined and lost the mea- sure of quantity, so did it lose also the notes of accent so intimately concerned in it. Thus the acute degenerated into a mere stress, according to which we pronounce the Latin \ and which also in I We may here obBerve, that in Latin the acute falls on the penultimate of dissyllables, and of such words of more than two syllables as have the penultimate long. Where such words have it short, the accent falls on the antepenultimate. The accent on the last is little congenial with the language. 20 ON STRESS. the Greek still retains its old places, to the great disturbance of the quantity, as in ovXofiivr\v, which they are obliged to pronounce with the penultimate long, and in avdpwTrog which they must pronounce as a dactyl. For the stress on a single vowel necessarily prolongs its time, and the vowel on which it does not fall must therefore be of a shorter time. While on the principle of accent av might be pronounced in the note A, Opwirog in the note F, without the true quantity being at all affected. 14. It is impossible for us to conceive the rich- ness of the recitation of the old versification, rolling in varied measure of time to the music of its accents. The analogy between its basis and ours of stress has been well illustrated by the com- parison of the flute, which holds notes a longer or shorter time, with the drum, which expresses but louder or softer sounds ; and indeed the reading of a passage of Tasso or Milton after one of Homer or Virgil, has very much the same effect as the drums which are beaten after the full band has paused. The confusion between things so different has arisen from our being compelled to read the ancient measures according to stress, and in some degree also from the carelessness of applying the word Accent to our modem Stress. 15. The stress is as necessary to give clearness ON STRESS. 21 to speecli, as the division into words is to convey it to writing. It distinguishes a word from its neighbour, and would do so most perfectly, and our speech might run without a pause between the words, as the Greek is written in old inscriptions without a break, if the stress fell uniformly on the first, or uniformly on the last syllable of the word ; to the first of which conditions there is a prevailing inclination in our language, as to the last in the French. But the monotony would be so intoler- able, that we should pay dearly for superior pre- cision, and the stress would be wholly inapplicable as an instrument of verse, since it would be im- possible to secure the regular interval of a certain number of unaccented syllables, except at least through a short and painful extent. It falls, therefore, on either syllable of a dissyllable. But when we come to a trisyllable, and place it on the first, we find ourselves rather far back for marking off the whole word, and are inclined to repeat it, faintly indeed, on the last, as appears from our allowing such words as liberty, intemperate, to conclude such verses as necessarily end with an accented syllable. And here lies a great differ- ence between accent and stress. In accent there can be set a decided limit to the height of the voice. In the Greek it rises through the interval of a third. But in stress there can be no certain 22 ON STRESS. limit to loudness. Being thus undefinable, it ad- mits of at least two degrees, and we can so modify one stress in comparison of another, that it may either equal it, or be something less, or almost vanish, as it does at the end of these trisyllables. Thus, in proportion as we feel the need of it, the fainter and secondary stress becomes more sen- sible ; as it does when we place the principal stress so far back as on the fourth syllable from the end, for instance, in the words moderator, administra- tion. There we are obliged to make a very decided, if not so full a stress, on the next syllable but one after the place of the first stress : though indeed in many cases, as in the word recollection, it is diiScult to say which is the primary and which the secondary stress. Hence in our language the stress is echoed through every alternate syllable of a word, and our recitation has a waving motion, which becomes very striking in the recitation of what are called iambic or trochaic metres. Thus far it is an admirable instrument of versification, and such measures are almost exclusively the vehicles of our poetry. 16. Now, also, we see how we are to be supplied with what are called the dactylic and anapaestic measures, which require the stress but once in the space of three syllables. Though we cannot re- present these so perfectly as the last, yet we may ON STRESS. 23 exhibit them decidedly enough for the purpose. We have only to beep our ear alive to the stronger stress, and let the weaker go for nothing. And as the stress on the last syllable of such words as " liberty" is very faint, and as in a word of such length as " moderation," we may choose which we will for the stronger, the difficulty is overcome. For example, the word " moderation" will suit the iambic structure, as in To moderation give her due, or the dactylic, as in Give moderation her due. Again, the stress on the adjective may be weak- ened, compared with that of its following sub- stantive, so that such a combination as " the black deep," shall answer to an anapsest, though here the adjective should have a short syllable, as will be seen if we substitute wide for black ; the long syllable requiring a stress too strong to be suffi- ciently weakened. A similar arrangement may be made between the verb and its noun ; and in gene- ral, since in the case of two stresses coming together in one foot, one can be made weaker than the other, we can sufficiently annul its force, in com- parison with that of the other. StiU the metre will be clumsy and hobbling whenever we come to 24 ON STKESS. such an expedient, in proportion to the force of stress which we conventionally forego : and thus we see within what narrow hounds of harmony, stress, as compared with quantity, confines us. As a defective means of harmony, it has an analogy to the defective syntax of our modern tongues. They both compel us to exercise judg- ment and taste, and therefore to conventional usage, where the ancients had precluded such exercise by having already provided through it rigid rules of utterance, and definite forms of inflection. They could have known no such wide distinction between good and bad readers of poetry and prose as we are forced to admit, who must determine a dactyl from a cretic by the metre, and a nominative from an accusative from position or common sense. 17. There remains the consideration of the stress on monosyllables. It may at first sight seem necessary that the stress should fall on every monosyllable. But then we must bear in mind that there is in every language a class of words which contain no especial meaning in themselves, but only in relation to the neighbouring words, to which, therefore, we attach them with scarcely a sensible break in our pronunciation. Hence they can have no stress upon themselves, but add it instead to the stress of the word to which they ON STRESS. 25 are joined, as in "he w^nt," "give me/' "to t6wn," &c. Such are conjunctions, as " and, but, if," &c. ; pronouns, as " I, me, who," &c. ; articles, adverbs, prepositions ; auxiliary verbs, as " shall, will, am," &c. All such monosyllables are without stress, except when emphatic in sense ; thus, " I will go" may be pronounced "I will go," merely intimating the intention of going ; or " I' will go," signifying that none else shall go ; or " I will go," signifying a determination to go. When two or three such come together, then we must begin the stress with the most emphatic, and lay it on the rest so as not to leave more than two running without one, as in " But and if heC shall g6." 18. And here we see the remarkable advantage which our language possesses in having its auxi- liary verbs under the monosyllabic form ; for when they are unemphatic they slide unto the verb, and practically make one word with it, so that "he will d6" is almost as compact as " faciat." Thus they do not clog our sentences any thing like so much (to the ear at least, though not to the eye) as they do in French and Italian, where they are mostly dissyllabic. And in the next place, when they are emphatic, they convey a force of ex- pression which cannot be conveyed in Greek and Latin; for those languages must either employ c 26 ON STRESS. the same form to represent "he shall go/' and " he shall go," or vary the phrase. Nothing shows the scholarlike pronouncer of his own language more decidedly than the proper attention shown to the place of the stress, espe- cially in such clusters of unemphatic words as were specified above. Without it no one can have any sense of harmony, either in verse or prose : and yet how little is commonly shown, and how little instruction is bestowed on a point so neces- sary to good reading. The common habit of false emphasis arises out of this ignorance, which con- founds stress with emphasis. And surely no habit among those which are tolerated rouses one's pre- judice more quickly against the reader on the score of his proper requirements for a man of education. 27 CHAPTER IV. ON THE CONSTRUCTION OF A VERSE. 19. A VERSE is determined in quality by the order of the intervals and returns formerly mentioned, and in quantity by the number of them. 20. Each combination of interval and return, by which this order proceeds, is called a foot, being that on which the verse walks as it were. The interval may contain several syllables ; the return consists in general but of one. 21. The most simple combinations are of course dissyllabic, affording but one syllable for interval, as in Latin words, regnd, p&dens, and in the English, stately, remote. The latter of these feet is called in the classical tongues an iambus, the former a trochee. And the same terms are ap- plied in modern languages, though improperly, wherever stress takes the place of the long syl- lable, as in the above English words. 22. The next in order are trisyllabic, affording two syllables for the interval, as in the Latin word c2 28 ON THE CONSTBtJCTION sldera, and our English word harmony, and in tte Latin word segetes, and our English word maga- zine. The former are called dactyls, the latter anapaests. The above word (and it is not a native) is the only anapsest in one word in our language, and it is not very distinctly that. But we have the foot in quite sufficieht plenty as occupying part of one word, as in the word devastation, or parts of two or three words, as in the desire, to the woods. 23. The two ancient languages could (as we saw) have more than one return in one word, as in the Latin words Candidas, amlcos, mfausta; and some less common measures have the return ex- tending over two syllables together, as in the Bacchiac metre. This, however, is obviously im- possible in our languages which go by accent, and therefore such feet as the above words contain can be formed but by the junction of two or more words, as lovely child, a good use, bright glory. They are, however, seldom or never used in our English metres, although their eiFect may be good to a short extent. Such metres might be the fol- lowing : — Lovely child, sweet the smile on thy cheek. A good use is bad use, if m^n fail to ndte times. Fair daughter, why weepest ? (5 dry up these t^ar-drops. OF A VERSE. 29 One reason of their unfrequent use may be the indistinctness of many of these accents, according to what was said in Art. 16. Thus the last line may appear in a measure founded on the foot con- tained in the Latin iniqua, and in our ahwndant. 24!. The two ancient languages likewise used feet comprehending so much as even three returns, with one interval, as in their epitrite ; and three intervals with one return, as in their paeon ; besides the various combinations arising out of two inter- vals and two returns, as in the choriamb, Ionic, and antispast. But these are confined to their lyric poetry. Some of them are in our power also, as the choriamb, and antispast, and Ionic, as, Frdne in the dust, liest aldne, daughter of T^re ? wearied with groan. The lark, soaring amid glory, in heaven singeth her loud story. Bright Seraphim, sing merrily, shout jdyfully. In the bright b^am the sublime h6st to our God sings. And great has been the mistake of our lyric poets in not adopting measures peculiar to their depart- ment, corresponding to these. The Grermans have managed better. Such feet as combine two or more returns in immediate consequence can of course be effected in our language, and in any versifying upon stress, only by hobbling through c 3 so ON THE CONSTKUCTION one or more monosyllabic verts or nouns as above, and as tbe first epitrite is formed by the hnght sun sMnes. On the contrary, the pseons abound in our longer words, as in moderation, and a measure may be readily constructed upon them, as. Liberty abundantly rejoiceth, and immoderately. But it would clearly be a work of mere pedantry. 25. These compound feet are uncommon, and the grand staple of metre in all the European lan- guages is made up of the dissyllabic feet in which the proportion of interval and return is in quan- tity 1 : 2 or 2 : 1, as in the iambus and trochee ; and of the trisyllabic feet, where these are equal, as in the anapsest and dactyl As to accent, the case is reversed ; in the former the proportion of syllables is equal, in the latter in the proportion of 2 : 1 or 1 : 2. These are the simplest propor- tions, such as the ear readily catches. Other feet, like the cretic and amphibrach, which admit of no simpler division than in the proportion of 2 : 3, or its alternate, are too elaborate for common use. 26. The rule of quantity has one very especial advantage over that of stress. For it is obvious, that since the interval depends upon the quan- tity, and not on the number of its syllables, it will make no difference whether it be made up of one or two, or of two and four syllables. OF A VERSE. 31 SO long as tlie sense of the return is maintained with sufficient distinctness. Hence the two short of the dactyl and anapaest may be replaced by a long syllable ; that is, a spondee may be sub- stituted for those feet, provided that the pure original foot recur often enough to remind the ear of the cadence. In the highly elaborate opening of the Georgics the dactyl is found to recur quite sufficiently often in the four first feet (where alone substitution is commonly allowed), though it does but in the proportion of 37 : 42. Hence also the tribrach may be substituted for the iambus and trochee. 27. But since the sense of the return depends upon one syllable, it is equally obvious that we cannot substitute two short for that one long, un- less we preserve the proportions of the interval and return. Thus we can substitute the dactyl for the anapaest, because the return divides them equally (,^ v-- -ij _ v^ w) ; but we cannot substitute the anapaest for the dactyl, because then the divi- sion is not only unequal (-i ^^ v^, vS «_» — ), but out of all harmonious proportion. So also the division is unequal between the iambus and trocheOj and they can never be mutual substitutes. Still more, the amphibrach cannot stand for either dactyl i-^^yj> 'L-^) °^ anapaest (^ ^ JL, ^ j. ^j), be- cause it is impossible to divide it in the proportion c 4 32 ON THE CONSTKUCTION of either interval or return. But the tribrach, ^L> \jr 3 59 „ 45 17 >» 1 58 >» 43 22 »» 54 13 >i 1 448 464 Thus there are, as nearly as can be, nine pauses to two lines : and the average number of pauses in the part of the line preceding the closing dactyl and spondee will be rather more than three ; three, however, is the more regular number, as we might expect, since an extent of from eight to twelve D 3 54 ON THE PAUSES syllables are best divided thus, according to the common length of Greek and Latin words. Thus the model is, Extinctiuii I nymphse | crudeli | ftinere Daphnim. — Bel. v. 20. Pulverulenta | coquat | maturis | solibus sestas. — Georg. i. 66. At least in the Virgilian hexameter : and how- ever partially broken by inferior pauses, this arrangement prevails in the main, as in Georg. i. 34. 29 :— Panditur | ipse tibi | jam brachia | contrahit ardens. An Deus ( immensi | venjas maris | ac tua nautse. 47. There are some circumstances attending these pauses which require notice. In the first place, since the sense of the grammar and of the metre should not be found contradictory, the stops of the sentence and the pauses of the metre should be of the same relative strength. The stops, therefore, will generally be faint in the beginning, and increase in strength towards the middle, where they become fullest, and thence will decline again towards the end of the line. Of course this rule is not kept with severe exactness. But where it is palpably violated, some good reason is seen why it should have so been. 48. To take them now in order, as to the pause after the long syllable ; it is of course very faint OF THE HEXAMETER. 55 after the first, so that the monosyllable com- mencing a line is almost always intimately con- nected with the following word without a stop, as in Gens inimica, Fert animus : or if there be a stop of any fulness, the abruptness is subdued by the recitation being carried immediately on to the next word by elision, as in those instances pro- duced by Hermann (El. Doctr. Metr. 2. xxvi.) from Iliad, K 51, Odyss. M. BdW. 'Aei Si Trupat, k.j-.X. "Oip'. 'H/ioe S' kiTi Sopirov, k.t.\. But the above list shows that this pause is rather unfrequent altogether. When we come to the third and fourth long syllables, we arrive at the caesuras of the line, as it is constructed by Virgil and the Latin poets. They are very strongly marked, falling each close to the middle of the line, and a change of measure to anapaestic ensuing from them. Hence these are the favourite places for the full stop, as in ^n. i. 76. 110 :— ^olua htec contra. | Tuns, O re^na, qiiid optes. Dorsum immane mari summo. | Tres Eurus ab alto. After the fifth long syllable the pause is infi-e- quent even in Homer, and scarcely used by Virgil except in case of a proper name, as in Mn. iii. 401, Lyctius Idomeneus : hie ilia ducis | Meliboei. D 4 56 ON THE PAUSES Otherwise he breaks its strength by a neighbour- ing pause of greater strength, as in Georg. i. 80 : — Ne saturare fimo pingui pudeat | sola | neve. Once or twice indeed Virgil copies the familiar pause of ni}Xj)io8Ew | 'Ax'X^oe somewhat pedan- tically. And since this is far removed from the influence of any other pause on either side, it becomes so marked, that a vowel will stand before it without elision, as in the Homeric example. Thus we have Eel. x. 12, Aonia \ Aganippe, and in jEn. ix. 477, Fcemineo | ululatu. And here we may remark, that where this long syllable stands without elision, or is produced, being short, it is commonly preceded by a dactyl, as in Georg. i. 4. — Sit pecori | apibus. .^n. iv. 64. — Pectoribus | inhians 222. — Tunc sic Mercurium adloqviitur | ac — Perhaps the reason is, that the dactyl stands in greater opposition to the anapaestic measure which follows this pause ; and especially where an ana- paest also follows it immediately, as in the two former instances of the above three lines. There the pause is nearly as strong as the close of a line, and will bear the same licence. The line Ter sunt conati | imponere Pelio Ossam is a designed violation of this more common rule. OF THE HEXAMETER. 57 We should not expect a stop after the sixth long syllable, nor does it occur in Virgil, except after such phrases as " hominum | rex," or as "Humi I bos, exiguus | mus, magnis | Dis,"' where the effect sought is clear, and the pause is but weak. In Eel. vii. 35, — Nunc te marmoreum pro tempore fecimns. At | tn, — the very weak, if any, pause after the sixth long syllable, is quite absorbed in the full stop at the end of the fifth foot. The pause after the long syllable gives much steadiness and stateliness to the line, which thus marches as it were: it is characteristic of the Latin. 49. But the trochaic pause gives a light step to the measure, and is as characteristic of the Greek. It is not, however, common until we come to the third foot : there it is a cassura with Homer, and imparts its lively influence to the line ; while it is unfrequent with Virgil. But both poets have a great reluctance to the use of the fourth trochee. We seldom meet with such a line as ^n. i. 77, Explorare labor. Mihi jussa | capessere fas est ; where, however the strength of the preceding D 5 58 ON THE PAUSES caesura weakens its effect. We quite lose sight of it in such a line as, Mn. i. 10^ Insignem pietate virum | tot | adiie labores. After the fifth trochee both poets have a favourite rest, Homer more especially ; while Virgil dwells more on the rest at the end of the fifth foot, which Homer uses comparatively seldom : both poets thus maintaining their cha- racteristic step. 50. The pause after the foot divides the line into its units, and therefore requires especial caution towards the middle of the line. After the first foot it is very agreeable, allowing us to start with a full sense of the measure, especially when it follows a dactyl. Also at the end of the fourth foot its good effect is felt by its marking off the regular close of the line from the rest ; and at the end of the fifth of course, distinctly marking off the dactyl, it is in its place. But after the second, and especially the third, foot, where it breaks the line exactly in two, it hardly occurs, except when weakened and made insensible by a neighbouring pause. Thus Virgil would not begin a line with conticuerunt. But it is in- sensible in Vina bonus | qute | deinde. OF THE HEXAMETER. 59 The common method of deadeniqg the pause in these two places with VirgU, is to put a pyrrhic dissyllable before the pause, which enfeebles it by the close neighbourhood of another, as in ^n. ii. 2. 776. Inde toro | pater | iEneas sic orsus ab alto. Quid tantum insano { juvat | indulgere dolori. In the former of these cases, he almost always places a pause again in the middle of the fourth foot, as above. And whenever he adds a stop to the pause, he is wont to break its force by putting a monosyllable after it, as in Mn. i. 52, iGoliam | veuit. | Hie | vasto rex .£olus antro. In the latter case, he very often combines a stop with the pause which precedes or follows, and thus most effectually breaks its force, as in Mn. ii. 27. 354, Panduntur portse : | juvat | ire, et Dorica castra. Sic animis juvenum | furor | additus. Inde lupi cen. Sometimes he will even combine a stop with the pause itself, but then he breaks it, as in ^n. ii. 528. Georg. i. 358. Porticibus longis | fugit, | et vacua atria lustrat. Montibus audiri | fragor ; | aut resonantia longe. And not only does Virgil terminate the foot in a D 6 60 ON THE PAUSES pyrrhic dissyllable in these places; but very frequently in others also, as in Georg. i. 363. — Deserit atque altam supra | volat | ardea nubem. 356. — Continuo ventis surgentibus aut | freta | ponti. In which last line, however, there is hardly a sensible pause between " aut" and " freta." The pause at the end of the fourth foot is a favourite with the Greek, and especially with the Bucolic, poets, who employ the dactyl before it. The Odyssey opens with a line containing it : 'AvSpa juoi tvviTTi, Movaa, TroXirpoTrov. | 8s iia\a iroXXd. Here the full stop is very frequent in the Greek. A stop of any strength at the end of the fifth foot does not agree with the weakness of the pause there. It seldom therefore occurs ; and then its force is broken by the following word being closely referred to the preceding, as in ^n. X. 195. — Ingentem remis Centaurum promovet. | lUe Georg. i. 80. — Ne saturare fimo pingui pudeat sola : | neve Eel. vii. 35. — Nunc te marmoreum pro tempore fecimus. [ At tu — 51. In their very nature these pauses afford, as we see, a most distinct and agreeable variety, and adapt the line to the manifold expression of thought and affection, and, above all, to the turns of narrative, through the quicker or slower move- ment which they impart to the line. But in their OF THE HEXAMETER. 61 number also they contribute to a variety whicb may be called inexhaustible. For suppose that there were but four pauses to a line : then, since the whole number admissible is sixteen, there may follow 1820 lines in continued succession, and yet each of different structure on this single point. The average, however, we have seen to be that of nine pauses to two lines. The series must therefore be a much longer one. Were the number of pauses five, the series would be as high as 4368. But when we combine with this the variety arising out of the permutations of the dactyl and spondee, there appears no practical limit to the changes which the measure can assume. Those permutations in the first four feet are sixteen. Taking the above series at 3000, we may then have 16 x 3000 consecutive lines of different metrical construction ; or, in other words, as much as two Iliads and two Odysseys together may be written without repetition of a single line of the very same metrical form. We indeed here suppose that these two causes of variety act always inde- pendently of each other, as they do not. Still we see enough to be struck with the wonderfiil power of this metre, and we must allow it to be the prince of measures, and admirably suited for narrative poetry. How do we sigh for its flexibi- lity, when we read the poor and monotonous 62 ON THE PAUSES strains with whicli modern epic poetry must be content. 52. Such then is this prodigious variety of structure, that it must be sheer negligence in- deed, nay even an effort of perverseness, when a line unites with the following in making a com- plete and correct verse between them. And how beautiful is the simplicity of the close ! It pre- sents but the sequence of a dactyl and spondee, which admits but of two varieties according to its more regular arrangement. These bring the ear distinctly to the close by their recurrence amid the unlimited variety of the rest of the line, at the same time that they do not cloy it with a monotonous ring. 53. We have had occasion to remark on the obvious difference of the structure of this measure by Homer and by Virgil, as to the arrangement of the pauses. This is the result of the difference of the character of the two languages, rather than of the difference of the genius of the two authors, which is after all but a particular case implied in the former difference. The Latin is more spondaic. And therefore, while in the Greek hexameter the recurrence of the dactyl in proportion to that of the spondee is as high as 25 : 11, in the Latin it is but as 9 : 11, or little more than a third of such proportion. Hence we cannot wonder that OF THE HEXAMETER. 63 the proportion of the number of trochaic pauses in Homer and in Virgil should be in the high ratio of 2 : 1. This very superior liveliness in the style of the narrative of the Greek has, without doubt, a large share in contributing to the feeling that Homer is telling us what he has seen, Virgil what he has read ; and is one of the means of that un- approachable superiority, on the whole, which he must ever maintain among the composers of epic poetry. 64 CHAPTER VII. ON THE PAUSES OF THE TKIMETEE. 54. NoTHiua displays the ricliness of the re- sources of the Greek tongue, as to the various and suitable means of expression, than the invention of two metres of such distinct character for two distinct departments of poetry, as the Hexameter and Trimeter. Here, as elsewhere, all the modem tongues exhibit poverty itself in the comparison. On going from one measure to the other we are immediately struck with the essential difference. It is indeed a descent ; for the trimeter is very inferior in its capability. It is constructed of a shorter foot, which however is admitted to the same number of six. And it allows of fewer pauses, and is deficient in variety and sonorousness of close. All, in short, warns us that we are come nearer to the province of prose, and from the reci- tation of the bard to the declamation of the actor — from narrative to conversation. 55. We find at once quite a different principle ON THE PAUSES OP THE TRIMETER. 65 of construction. The spondee is introduced not as an exact equivalent in time, but as an avowedly different foot called in to modify the pace of the line, and impart stability to its undulating move- ment. And this foot brings in also its equivalents under certain regulations, so that this dissyllabic measure allows of three trisyllabic feet, the tri- brach as the equivalent of the iambus, the dactyl and anapaest as equivalents of the spondee. A great variety is thus obtained. But since it arises from foreign help, and not out of a development of the measure itself, it wants both the extent, and force, and beauty of that which the hexameter affords. 56. The introduction of these feet must of course be kept within such bounds that it shall not de- range the line. And therefore, where greater care was bestowed, as by the two earlier tragedians, who maintained a higher tone than the third, pro- vision was made that those essential features, the pauses of the line, should not be affected. There- fore the two short syllables, which, on the reso- lution of the spondee, take place of the one short of the iambus, were not allowed to interfere with their regulation, and no pause was admitted after either one or both, but the anapaest to which they belonged was contained in one word, such as fisyaXriv, not in two, as in rpie | iiftira, fxiya | 66 ON THE PAUSES Tovro ; the former of which gives an iambic instead of trochaic movement after the pause, as in that place it should : the latter comes after two short syllables instead of one, and so quite deranges the proper pause. Other ways, in which these sub- stitutions are restricted, wUl appear when we come to consider the pauses in particular. 57. The number of pauses is altogether eleven: for six can fall in the middle of the foot ; after each of these the movement becomes trochaic : and five can fall at the end of a foot ; after which the movement still continues iambic. The follow- ing scheme will show the comparative frequency of the occurrence of these pauses within the range of ] 00 lines of Sophocles. Foot I. In middle of first foot. . 21 At end of ditto .... 35 II. In middle 50 At end 11 III. In middle 80 At end 6 IV. In middle 60 At end 31 V. In middle 18 At end 52 VI. In middle 2 Sum 366 Or 3.66 to a line, i. e. about 7 to two lines. 58. In the first foot the more regular construc- tion restrained not only the anapaest, but the OP THE TRIMETER. 67 dactyl also within a word. It is obvious that it never could be allowed to be broken into a trochee followed by a short syllable, as ravra | jusXetSts, since such a pause is inadmissible at the beginning of an iambic line, uttering the very reverse of its measure at starting. And there would be a re- luctance to inclose in any place a pyrrhic within two pauses, since they shoidd, on the principles of the verse, inclose but one short syllable. Hence a dactyl distributed between two words, the first of which was a long syllable, such as ^ iroaiv, was avoided. But in fact the dactyl is rare at all in the first place in the more regular measure. Nor can we wonder, since it runs in the contrary direc- tion, that of the trochee, and is therefore not very suitable for opening an iambic movement. Hence the anapaest is much more common in this place. A favourite pause falls in the middle of the second foot. This pushes the line as it were into a brisk movement as early as a sense of the tro- chaic pace could be borne. Compare the sixteenth line with the seventeenth of the CEdipus Tyrannus. At the end it is rare, on account of the close neighbourhood of the principal esesura. As in the Virgilian hexameter, we find the prin- cipal pauses, that is, caesuras, in the middle of the third and fourth feet : supplying also a similar variety by change of foot to the opposite, the 68 ON THE PAUSES iambic becoming trocbaic, on the recitation starting from them. The former of these places presents that which is the favourite, and predominates in the proportion of 4 : 3. Many lines employ both, as in (Ed. T. 7 :— 'ANXuv uKoiitv I oiSrJj | US' i\ri\v9a. This former pause, the penthimimeral, as it is called, is a main reason for rejecting the anapaest as a substitute for the spondee in the third foot ; for it must divide it in one of the following ways, 'Ajrpaicra ravra | Xly£i 'ESe^d/iriv jiiya \ SiSpov 'ErXriaa itofKrifiov | dx^oQ, all of which are objectionable on the principle laid down in Art. 56. Besides that the mind, being so accustomed through this pause as to feel its place, even where it does not occur, will not tolerate two syllables instead of one in it, and therefore rejects also such a line as Kai /iriv avcv BepdxovTos \ iitiXBiv So/iiov. Thus, under no circumstances, can the anapaest stand in the third foot. But the same foot is the favourite place of the dactyl, which is so introduced, that the penthi- mimeral pause falls after its long syllable, and then the trochaic movement commences with a OF THE TEIMETER. 69 tribrach ; which, however, on the principle above laid down, should not commence with a dissyllable, at least on the principle of construction followed by ^schylus, Sophocles, and Euripides in his earlier plays. Similarly the favourite position of the tribrach in the fourth place is that which places its first syllable just before the hepthi- mimerd caesura, and the remaining in a word which more than contains the rest ; and the fifth foot is most commonly in this case pure, as in Orest. 24, 'Xporfv T 'OpeffrjjSi /iriTpis | avoaiutTaTtiQ, Less regular is ib. 74, TXilliiDV 'GpeoTijg juijrpof 'oSt ^owig ejiv. It is comparatively rare where this caesura does not occur to break the tribrach, as in ib. 60, npouTTe/tipEJ' is Sdii riairtpov lanv S' laut. Add (Ed. T. 960. But not even Euripides, still less ^schylus, could have written such a line as has been palmed on him in Pr. V. S62, TvipSiva 9ovpov, iraaiv 8g aviary BeoiQ. Sophocles seldom uses it at all, except where a proper name occurs. So rare is the bisecting pause, that some have treated the lines which present it as corrupt. We 70 ON THE PAUSES have, however, seen reason for its use, for the sake of peculiar expression; and something must he allowed in the way of indulgence, or on the plea of variety of expression, to the poet, as in 'AXX' Sv TToXie arvyu \ ail Tt/iijaiiQ vsKpov. Most commonly it takes place after elision, as in (Ed. T. 46, so as to partake of the nature of the hepthimimeral. It sometimes becomes insen^ sible, from the close neighbourhood of another pause, as in (Ed. T. 77, M^ Sjiiuv av Eiiji' I iravB', | oa av ItjXoI Stofi. The pause at the end of the fourth, leaves the rest of the line like the commencement of a fresh line, and thus causing it to seem to overflow into the next, gives a freedom which brings the passage nearer to the tone of common dialogue. It does not, however, occur so often as might be expected, not only because the tragic poet sought to main- tain some elevation, but because its combination with the same pause in the preceding or following lines included a verse, however internally incorrect, and produced thus more vagueness than could be allowed. For the same reason the pause at the end of the second foot is rare. The pause at the end of the fifth foot is very common. It serves to mark off the close by OF THE TRIMETER. 71 leaving an iambus in the final word : and hence also the infrequency of the pause in the middle of this foot, which leaves a cretic in the final word. In this case the effect was generally broken by a pause at the end of the fourth, which brings out an iambic close, as in Toil \6yov | tovt' [ l^fpw. Or they made the iambic movement clear, by keeping the fifth foot an iambus, as in Kal TT€(76vTeQ I uartpov, So that they could not admit such terminations as ToiiTOVQ I iKepH- TTitrovTuiv I iiffrepov. The fifth foot comes so near to the close that great caution is required, so much so, that even the tribrach was admitted but under very rare and peculiar qualification. Of course, then, tlie dactyl was avoided, which, in addition to the fault of the tribrach which confuses the sense of the final iambus by running into its short syllable with two short, has others of its own, as having only indirectly, through the licence of the spondee, a place in the line. The anapaest also is quite inadmissible, when divided between two words, were it only on the principle which excluded it 72 ON THE PAUSES from tlie third place : and when contained in a trisyllable, it still disturbs the regular trochaic flow of the line, which is felt on quitting the place of the first caesura. The close is faint indeed, compared with that of the hexameter, but quite accords with the more quiet pace of the iambic rhythm, which came so near to that of common speech : and when even such strictness as this had was relaxed by in- troducing the spondee and its equivalents, the measure was admirably suited to the purposes of the tragic and comic poets ; the latter, indeed, by their further licences, in admission of feet and neglect of pauses, brought it down as near to prose as possible. ^ 59. The number of pauses admissible being eleven, and the number admitted being usually four, or rather seven to two lines, there may follow in series about as many as 330 lines of different construction on this point. This is hardly a tenth part of the number afforded by the hexameter: it admits, however, of more variety still, from the variety of its constituent feet, and will, on this account only, allow of 240 changes. Hence alto- gether 79,200 lines of different construction may follow in series, or more than twice as many Greek plays as exist may be written without the repeti- tion of the same construction of the line, on the OP THE TEIMETEK. 73 same supposition as was made in the case of the hexameter. So large is the scope, and free is the compass, within which the Greek constructed its two grand measures for epic and tragic poetry, which are each as different from the other in character, and peculiarly adapted to their subjects, as language seems able to allow. 60. The very feet of those measures, namely, the dactyl and iambus, are of quite a different character. The first has a roU forwards, the latter a fling backwards : the first seems fitted to carry the imagination on its course, the second the reflection; the former adapted to epic majesty and didactic description, the latter to moral sen- tentiousness and returns of dialogue. Through use of the spondee, the hexameter can unfold or draw up its ample robes, like an Ionian matron, with most imposing dignity, now swelling up to seventeen syllables, then shrinking to thirteen, as when we compare Quadrupedante pedum sonitu quatit ungula campum, with OIU inter sese magni vi brachia toUunt. While through the same use, though that be not in the legitimate way of an equivale'>t, the tri- meter, like a Doric girl, can give now and then a flirt or two to her kirtle through the extent of from 74 ON THE PAUSES twelve to fifteen syllables (in the pure tragic form), as on comparing (Ed. T. 966, 967, Kraveiv IfieWov Traripa rbv i/wv ; 6 Sk daviav. Perhaps this difference of character could not be more forcibly exhibited than by the juxta- position in which Horace has placed these mea- sures in the sixteenth of his Epodes. These both are remarkably regular, the trimeter consisting of pure iambi, and having invariably the caesura in the middle of the third, which is also mostly the case with the hexameter : thus they are brought together in their essential forms, so that we can discern most clearly their distinctive differences. What a remarkable change do we experience when the majestic roll of the hexameter, with its sonorous close, is succeeded by the uniform alter- nate step of the trimeter, with its faint termi- nation ! It is greater than that between marching and walking. The same effect is evident, when we compare the description of an event given by Homer, and by a tragic poet in the mouth of a messenger, especially in the play of the Persse, or Seven against Thebes, which -come so completely on the same ground with the epic poet. With all their liveliness of description, what a prosaic poverty do those passages exhibit compared with OF THE TRIMETER. 75 any similar in Homer. The sense of the manner in which his measure would have handled them haunts us as we read. We are looking at a meeting-house with our head fuU of a church, or hearing its service with our ears yet ringing with the rich tones of our chanted liturgy. But if the trimeter he so inferior, what are we to think of the one measure with which modem epic and tragedy must both needs be content, shorter as it is by a foot, and regulated by mere stress. It does, indeed, appear poverty itself. Nothing but a resolute forgetfulness of the ancient measures can make us feel any satisfaction with it. Take up the very best translation out of the many which have been made of fine passages in Homer or Virgil, after having read the original, and then how lamentable appears the inadequacy of the copy ! Or compare a cleverly translated passage of Milton into Greek hexameter, or of Shakespeare into Greek trimeter, and then how wonderfully improved do they seem ! E 2 76 CHAPTER VIII. ON THE PAUSES OF THE ALEXANDRINE. 61. The name " Alexandrine" is applied in the French and in the English to two measures, which have no other resemblance than that they consist of the same number of clear syllables, though not of feet, and have uniformly the bisecting pause ; the foot being in the French the anapsest, in the English the iambus. Yet notwithstanding the coincidence of the latter with the Greek trimeter, it seems so little to have been founded upon it, that its history shows that it is nothing more than the former reduced into correspondence with the rhythm of the language, which is iambic, as that of the French is anapaestic : and for the origin of the former we are on every ground of probability referred to the dactylic pentameter, as pronounced according to accent, as Thebani media non sine matre duces. the verse containing the feminine rhyme being considered the primitive type. ON THE PAUSES OE THE ALEXANDBINE. 77 ' 62, The bisecting pause being so very strong as to blunt our sense almost to any other, it is not worth while to give any list of the pauses con- tained in this line, especially since in our language the measure, considered as one verse, has long ago been obsolete : and where it is in^eijected, as at great intervals it sometimes is, among our ten- syllabled iambic lines, since it then comes single, the comparison of the pauses of two consecutive lines on which the sense of their harmony so much depends, is lost. It needs but to be ob- served, that from the monosyllabic nature of our language, the two pauses which are next in strength, coming each one syllable before and one after this bisecting pause, can seldom both of them be avoided; and the first comes continually to- gether with it, and gives a most monotonous ring at the middle of the verse, from line to line, their csesura falling on a monosyllable, as in "Poly- olbion," X. 1, Awhile thus taking | breath, | our way yet fair in view, The Muse her former | course | doth seriously pursue, From Penmaen's craggy | height | to try her saily wings, Herself long having | bathed | in the delicious springs. Thus our Alexandrine is even more intolerable than the French, exceeding even that in the dull repe- tition of equal parts, as if the evil genius of versi- fication had not already invented enough. It is E 3 78 ON THE PAUSES truly in the garden of poetry that which Pope ridiculed in the gardens, of his day. Grove nods at grove, each alley has its brother : And half the platform just reflects the other. We have seen, indeed, that Spenser occasionally violates this rule of uniformity (40) ; but dared he have so done in a series of lines, such as Drayton uses. Amid such rigour the ear could hardly endure the licence. Happily, therefore, the "Polyolbion" is the latest poem which our language affords constructed on this measure, although not the only poem ; for the measure is as ancient in our language as the thirteenth century, in which lived Robert of Gloucester, whose his- torical poem on England might have led Drayton, antiquarian as he was, to adopt the same metre for his topographical poem on the same. So unfit, however, was it henceforward found, that it was not even revived among foreign importations at the period of the general debasement of our poetry by French imitators, under the auspices of the Frenchified Charles II., half a century later. 63. Even so strong a close as rhyme has not pTOved sufficient to prevent that running of one line into another, to which we have seen this measure to be peculiarly exposed (43) : a further remedy, therefore, was applied by the French, OF THE ALEXANDRINE. 79 faamely, the closing the sense with the line, and which, indeed, the nature of rhyme, strictly followed, requires. But this has naturally given rise to that antithetic cast and epigrammatic point, which so much pervades French poetry. One hemistich is balanced in sense, as well as in measure, against another, sometimes giving out a variation of the sense, like the clauses of Hehrew .poetry, and sometimes charged with a heap of epithets belong- ing to a noun which is contained in the former hemistich, or with an apposition of nouns governed by verbs in the former. So far the measure is not unsuitable to the rhetorical flourish and de- clamatory strain which distinguishes the French theatre : but for every other purpose, never was a measure so ill suited. It must, however, be allowed to have one great advantage in French verse which it has not in ours, and that is, the alternate mixture of masculine and feminine rhymes. This to a more certain degree prevents the confusion of one line with another. 64. When a line has been extended to a certain length, the mind and the breath are fain to make the principal caesura as marked as the close. Hence the tendency of such measures to break into two. The limit at which this dismember- ment takes place, depends on the character of th language. If its words be short, then, since the E 4 80 ON THE PAUSES quantity of sense to a line is limited, from the mind requiring a certain compass of meaning, no less than the ear of sound, its metrical lines will be shorter, and therefore will break sooner when stretched beyond the usual point. Thus in our tongue the words are like the stones in our build- ings — small, and our structure of them breaks into small masses ; while in the Greek and Latin they are like their marble blocks, which allow of large masses. Hence not only do the pauses in our verses coincide too much with the end of the feet ; but also our longer measures cannot sustain them- selves, but break asunder: and our Alexandrine now survives only in the lyric stanza of four lines and rhymes, such as occurs in Psalm cxlviii 1 : Ye boundless realms of joy, Exalt your Maker's fame : His praise your song employ, Above the starry frame. But not the slightest tendency to such a breach is observable in the Greek trimeter ; and that sonorous tongue can sustain tetrameters where we have never even ventured, as in the trochaic and anapaestic ; or ventured in vain, as in the iambic, employed by Chapman. 65. Such a fact demonstrates that this measure has too lyrical a cast for our modem narrative OF THE ALEXANDRINE. 81 poetry. And yet, at the same time, it is too poor for lyrical poetry of a high order. We can only wonder at the taste which could endure any long continuation of it, whether used in its entire or broken form. Its marked caesura and close make it more lyrical than the hexameter: the insig- nificance of its intermediate parts more prosaic than the trimeter. Hence its real character is that of mere hallad ; and we may dismiss it from any further consideration upon the staple measures of versification. E 5 CHAPTER IX. ON THE PAUSES OF THE HEROIC. 66. The heroic line of ten iambic feet, borrowed by us througb Chaucer, if through none earlier, from the French, appears to owe its origin to the Latin hendecasyllable. This is more discernible in the Italian form of it, which consists of eleven syllables, and freely admits the trochee ; as, for example, in the very first line of the " JelTisalem Delivered : " Canto I'armi pietosi e'l Capitano : the measure of which, according to the analogy of quantity, corresponds with Aiida modo pumice expolitum, of stress, with Dextram st^rnuit ipprobatidnuin ; while the common run of the line is according to the stress in Ad coe'lum lepido Tocare versu. ON THE PAUSES OF THE HEROIC. 83 On coming into the .bands of the French, it lost the last syllable in the masculine rhymes; and in this form, agreeably with the monosyllabic character of our language, was imported into English poetry, and, as early as the days of Chaucer, became its staple measure for epic and didactic subjects. It does indeed with us occa- sionally resume the double ending; but hardly ever, in the case of rhyme, unless a ludicrous effect be intended ; as in WMch, now to sense, and noyr to nonsense leaning, Means not, but blunders round about a meaning. Even without rhyme this is rare, — occurring but now and then in the " Paradise Lost," as in Of rebel angels, by whose aid aspiring — except in dramatic poetry, which is principally distinguished from any other by its free use of it, and even prolonging it to another syllable, still so as to end the line with a dactyl, in the place of the long syllable. For example : Of the Imperial theme. I thank you, | gentlemen. My thought, whose murder yet is but fan|tastical. Macbeth, i. 3. Otway's lines abound in such terminations. E 6 84 ON THE PAUSES 67. The number and frequency of the pauses in the course of a hundred lines is as follows : — I. After 1st syllable, in Tasso, 19, in Milton, 21, in Dryden, 20 61 28 „ 68 „ ,, 19 „ „ 72 „ II 15 ti 72 „ 20 . . 374 „ II. 2nd III. 3rd IV. 4tli V. Sib VI. eth Vll. 7tli VIII. 8th IX. 9th Sum 47 49 36 39 41 60 43 28 41 62 40 24 50 50 30 37 49 369 Hence the number of pauses is about seven in two lines. But taking it as high as four in a verse, we may then have 126 lines of different construction in succession ; and thus the measure, on this score, admits of not much more than a third of the variety of the trimeter, nor than a thirtieth of that of the hexameter. How unfit must it thus be for translating out of the latter ! "We can hardly recognize the original poet under such a metamorphosis. Homer seems to us trans- formed from the old bard in flowing and em- broidered robes, to the schoolboy in close jacket and trousers all dark grey : and we wonder what has become of the animation and variety which should distinguish the higher narrative poetry. OF THE HEROIC. 85 In order to adopt it to such a purpose, it must have these infused from some external source. Hence some epic poets, as Dante, Tasso, Spenser, have had recourse to stanzaic arrangement. How far they have overcome the difficulty, we shall see presently, when we proceed to compare the re- spective merits of the several measures according to the service to which they have been ordinarily applied. 68. We cannot but be struck on this view with the comparative equality of the recurrence of the pauses in our language, especially as employed by Milton. This arises from the monosyllabic nature of our language, which obliges us to resolve the line so much into its elementary feet, even down to syllables. Hence the cause of much monotony in our recitation ; which is very sensible to a foreign ear, as it listens to the declamation of our actors. But we must do Milton the justice to note his more frequent usage of the trochaic pauses, in spite of the obstacles interposed by our monosyllabic language ; while the Italian, with all his advantages, has appeared insensible to their value: so much so, that they are to the remainder hardly more than 1 : 3 in number ; and that ■which gives the greatest briskness of all, occurring after the fifth syllable, he has employed nearly the least of all. Hence there is a heaviness 86 ON THE PAUSES of roll in Ms lines, whicli, had he not it in common with Ariosto, we might have supposed to have been affected in imitation of what seemed to him a part of the gravity of Virgil ; it little accords with the animation of epic narrative, of which we have such an example in Homer. Nor, indeed, has Dryden by any means approached Milton in this particular. Compare them at the fifth and seventh pauses. The harmony of Dryden comes out more from the comparison of two lines, than from the consideration of one. 69. The csssuras are after the fourth and sixth syllable. But falling, as they do, at the end of feet, they produce no change of measure ; but only make it more marked. Hence too exclusive an use of them imparts a staid and strong, but monotonous character to the versification. And, moreover, it becomes requisite that great care should be taken in the management of them, that two lines containing the same caesura should not come together, since then they become con- founded one with another, as in This noble youth I to madness loved a dame Of high degree, I Honoria was her name. Theodore and Honoria, 10. Adom'd in ancient times I with arms and arts, And rich inhabitants I with generous hearts. — lb. 3. We have seen (43) how even Milton, the most OF THE HEROIC. careful constructor of verse, has fallen into this error. In Thomson, the most careless and ig- norant of all, we may sometimes draw a line down through almost a page, marking off the former of these csssuras in each line. And indeed so nearly does it, in its strength, resemble the end of a line, that in the dramatic poets we often find the same licence as occurs at the end of a line, namely, of a double termination ; as in " Macbeth," i. 6 :— To plague the in|ventor, | This evenhanded justice. But the stop is commonly full, as here. If not, we may refer it to the licence of the anapaest for the iambus (30). These caesuras divide the line in most harmo- nious proportions, those of 2 : 3, and 3:2. At the same time, the bisecting caesura makes amends for its uniformity of division by the change of measure which ensues upon it, as in " Theodore and Honoria," i. : — Of all the cities | in Romanian lands. Thus this measure has rich resources of good har- mony on this head, though they be not so various as in those which we have before reviewed. 70. When we come to the effect of the pauses on the measure, we are immediately struck with a strange contrast to ancient usage in the admission 88 ON THE PAUSES « of the trochee. This is a remnant of the old con- struction of the line, as we have already seen (66), and is admitted by the Italian into every foot but the last, and even two are allowed to stand together, as in the very first line of Tasso's epic. Milton, who was so deeply imbued with Italian poetry, and had a pedantic turn, accordingly admits it indis- criminately, as in "Paradise Lost :" — Shoots invisible virtue e'en to the deep. — iii. 587- In the visions of God. It was a hill. — xi. 377- Thy lingering, or vrith one stroke of his dart. — Jb, The English construction of the line, however, commonly rejects the foot, except in the first place. Pope is fond of it here. It gives a brisk start to the verse, as in "Essay on Man," i. : — Warms in the sun, refreshes in the breeze, Glows In the stars, and blossoms in the trees. But there is commonly a pause, either in the first foot, as in the above lines, or after it, as in Dryden, " Palamon and Arcite :" — Serious | in aspect, earnest in their talk. Such a line as Poverty | breaketh noble hearts in twain, however common in dramatic poetry, will hardly be found in the school of Dryden and Pope. OF THE HEROIC. 89 Once admitted in the first place, it would gain admission into the places following the csesuras (48), where however it seldom occurs except in older writers and dramatic poets. 71. Admitting the above licence, as also that of the double and triple termination, and that of trisyllabic feet in any part of the line, dramatic writers have formed a characteristic measure, dif- fering from that of the heroic much in the same way as the comic trimeter differed from the tragic. It may be brought thus to the very verge of prose, as we may see from the arrangement of parts of "Spectator," No. 459:— For instance, in that disputable point, Of persecuting men for conscience' sake. Besides embittering their minds with hatred. * It: * * * Distress their fortunes, hurt then- reputations. Ruin their families, make their lives painfiil, Or put an end to them. Sure when I see — ***** I'd be as foUy convinced o' th' truth of it As of a mathematical demonstration. It is to be regretted, therefore, that our comic poets should have composed in prose rather than in this loose metre. The difference of trouble to the author would have been slight, while that of lite- rary value is incalculable. He would communicate an ideal cast, which never should be wanting, even 90 ON THE PAUSES OF THE HBEOIC. to the most familiar representation of human life, be it even the joke of a clown. At the same time also we lament to find how inferior is such a strain to that which the Greek comic metre breathes,, maintaining the poetical and ideal cast of the tragic down to the lowest relaxation of its form. It can never fall into prose, nor can prose ever rise into it. But it maintains a character in exact harmony with its subject, apd with nothing else. Indeed this measure of ours and the trimeter may be considered as no very inaccurate exponents of the nature of the two theatres, which is as different as the statues of Apollo and Silenus are from those of Newton and Sutor John. Such considerations only illustrate still more its defectiveness as a . measure for heroic poetry, without borrowing help from such an extrinsical source as stanzaic arrange- ment. Its highest pitch should not carry it higher than moral poetry and conversation, beyond which even the trimeter did not aspire, except in the narratives of messengers in tragedy. But we are anticipating considerations which will be entered upon more at large presently. 91 CHAPTER X. ON THE PAUSES OF THE VARIOUS TETRAMETERS 72. We have seen that in proportion to the length of the line the caesuras become more marked, until they obtain all the force of closes, and the line breaks into two (64). The Greek, therefore, poly- syllabic though it was, and from that cause de- manding much more room for a sentence than English, and therefore also a longer range of verse, never pushed its metre beyond eight feet, or even quite so far, but stopped within a syllable or half a foot of it, as in the tetrameters composed of the iambus, of the trochee, of the anapaest. So marked are the caesuras in these, that it will be sufficient merely to quote examples in each : — 'Elriffxtf iv raXg daniaiv | Xo/S^v yap ivSiSioKas. Eqiiit. 844. 'Q Pa9vliibvtov avaaaa | IlipaiSwv vwtpTaTti. Pers. 158. Aeyc 9appfiaag' (ii£ rig anovdag | ov /t^ wpoTtpov irapa/Sujucv. Aves. 461. 92 ON THE PAUSES The other pauses, though not all equally constant, are sufficiently obvious after all that has been already stated, and therefore nothing remains for remark but the fact of these caesuras occurring after complete feet, so that the latter part of the line is an echo to the former, as we saw was the case in the Alexandrine (62). , It is, however, inevitable, from the peculiar length of the line, which naturally breaks at those points. Nor can we find fault with it, but on the contrary we feel how admirably such a monotonous repetition suits the uses to which the Greeks commonly put those measures, namely, to carrying on scenes peculiarly occupied with tragic agitation or comic bustle, which would be much heightened by language so strongly marked in intonation. But it should be observed, that no whole poems, but only parts in subordination, are composed in these measures, if we except the earliest tragedy, of which however no remains exist. Nothing shows more than this the nice discrimination and exquisite taste of the Greeks, whose genius gradually formed a language so flexible in all its departments, as to vary and give the proper expression to every mode of thought, as to whole and to parts. They can con- struct a building, in which the kitchen shows vnthout vulgarity that it is the kitchen, as much as the chapel shows with all solemnity that it is OF THE VARIOUS TETEAMETEKS. 93 the chapel. While our building, like too many- specimens of modern Gothic, can distinguish neither ; and, if it does not put a chimney to the chapel, puts a crocket to the kitchen. Pope's translation of the " Iliad" is an example, however extreme and partial, of this defect. 73. Some perhaps may think that these tetra- meters are only such to the eye, and are in reality broken into two parts, so that two tetrameters in sequel would form a stafP, just as in English, the second and fourth lines being shorter by a syllable than the first and third, as 'Q PaBvZiiviav avaaaa, ll€paiSlDV VTTipTaTI], M^«p r) !SiepKov ycpaia, Xalpe Aapeiou yvvai. It must be confessed that the csesura is so strong and constant in them all that we could not dis- tinguish it from a close, and must have allowed of such a disposition of lines, if tetrameters had always occurred in an even number, so as to pro- duce pairs, and the sense had almost always ter- minated with the end of each pair. But as this is far from being the case, there could have been no such stanzaic arrangement, and the lessons of the grammarians, which tell us that such tetrameters were whole verses to the ear as much as they are in print to our eyes, must be accepted. 92 ON THE PAUSES The other pauses, though not all equally constant, are sufficiently obvious after all that has been already stated, and therefore nothing remains for remark but the fact of these caesuras occurring after complete feet, so that the latter part of the line is an echo to the former, as we saw was the case in the Alexandrine (62). , It is, however, inevitable, from the peculiar length of the line, which naturally breaks at those points. Nor can we find fault with it, but on the contrary we feel how admirably such a monotonous repetition suits the uses to which the Greeks commonly put those measures, namely, to carrying on scenes peculiarly occupied with tragic agitation or comic bustle, which would be much heightened by language so strongly marked in intonation. But it should be observed, that no whole poems, but only parts in subordination, are composed in these measures, if we except the earliest tragedy, of which however no remains exist. Nothing shows more than this the nice discrimination and exquisite taste of the Greeks, whose genius gradually formed a language so flexible in all its departments, as to vary and give the proper expression to every mode of thought, as to whole and to parts. They can con- struct a building, in which the kitchen shows without vulgarity that it is the kitchen, as much as the chapel shows with all solemnity that it is OF THE VARIOUS TETRAMETERS. 93 the chapel. While our building, like too many specimens of modern Gothic, can distinguish neither ; and, if it does not put a chimney to the chapel, puts a crocket to the kitchen. Pope's translation of the " Iliad" is an example, however extreme and partial, of this defect. 73. Some perhaps may think that these tetra- meters are only such to the eye, and are in reality broken into two parts, so that two tetrameters in sequel would form a staff, just as in English, the second and fourth lines being shorter by a syllable than the first and third, as 'Q ^aBvZiivoiv avaaaa, JlepaiSiov WTrfprdrij, M^rep 4 !Siep^ov ytpaid, XaipE Aapciov yvi/ai. It must be confessed that the csesura is so strong and constant in them all that we could not dis- tinguish it from a close, and must have allowed of such a disposition of lines, if tetrameters had always occurred in an even number, so as to pro- duce pairs, and the sense had almost always ter- minated with the end of each pair. But as this is far from being the case, there could have been no such stanzaic arrangement, and the lessons of the grammarians, which tell us that such tetrameters were whole verses to the ear as much as they are in print to our eyes, must be accepted. 96 ON THE PAUSES, &C. it comes up to examine that whicli recommended itself at a distance. When the eye is confined within one of those divisions, it will not be con- tent with blank space, but must still have some lines, dimensions, and proportions to contemplate, be it but the junctures of the stones, or even of the bricks, as we must say in application to the words in our tongue, so short and plain as they are : therefore, as in verse each distinct word lies be- tween pauses, we come now to examine the cha- racter of the word itself, as it is concerned with recitation, and so we proceed to discuss such properties of syllables as contribute to the effect of verse. 97 CHAPTER XL ON THE EXPRESSION DUE TO THE AETICtlLATION OF SYLLABLES. 76. Considered under this head, syllables are of two classes: — (1) When nothing follows the vowel, as in 0, to, mu-sic. It is then called pure, or vocal. (2) When a consonant follows attached to the vowel, as in at, that, rude-ness, mus-ter. It is then called mixed, or consonantal. In the first of these cases, the accent, raising but the key, does not aiFect the quantity, as in d-tomus ' ; but the stress, by dwelling on the syllable, necessarily lengthens it, as in me-teor. In the second, the accent is equally ineffective : but the stress, including now the consonant also, dwells less on the vowel, and therefore its sound is not so full and long as before ; and the syllable is less sweet, though prolonged. Thus there is an ' We pronounce our word as at-om, not as a-tom. 98 ON THE EXPEESSION DUE increase of length, but also of closeness, and defi- ciency of fulness and sweetness in the progression, gray, grace, grac'd, or gray, graze, graz'd (9). This second class admits of subdivisions : — (1) When the attached consonant is a liquid, still more if it be compounded of one liquid fol- lowed by another, as in Arma, Arncliffe; or by a labial, as in Arva, Alhan, or even by a dental, Alahanda, Namancos : then the sound is full of majesty or sweetness, according to the nature of the vowel, as in Caernarvon. (2) If that consonant be a dental, and still more, a compound of two dentals, or of dental and guttural, as in haste, Atkinson, then the sound is always close, and often harsh. This latter cha- racter prevails through the Teutonic tongues, the former through the Celtic, as appears at once from comparing the names of places in England and Germany, with those in Wales, Ireland, and the Highlands, or even two names of one mountain. Saddleback, Blencaihro. (3) When the vowel is short, then the con- sonant, instead of being nearly absorbed in the long vowel, makes itself so sensible as to diminish much the sweetness. , Compare great and regret. Much, of course, depends here also on the nature of the consonant, which in the Teutonic tongues is too often a dental or a guttural. TO THE ARTICULATION OF SYLLABLES. 99 77. Before going into further particulars, it will be necessary to mark the degrees of these sounds. These are given in the following table, which shows at the same time the comparative preva- lence of each in the Greek, Latin, and English. It should be observed, that our w and y here go for consonants ; and the nearest sounds are put for the same, where these do not occur. I. Long. Greek, — . 1. cs or a in fall . . . < Latin, — . 4. i or e in feed 5. in four . . I English, 28. Greek, 67. 2. a or a in father . . -^ Latin, 125. English, 9. Greek, 137. 3. e or a in fate . . . -^ Latin, 147. English, 47. Greek, 22. English, Greek, 35. 6. u or 00 in food . . . -^ Latin, 85. English, 18. F 2 i r Greek, 22. < Latin, 108. t English, 74. r Greek, 89. ■< Latin, 67. (. English, 41. I 100 ON THE EXPKESSION DUE II. Shoet. i Greek, — . 7. a or in folly . . . -j Latin, — . ( English, 84 f Greek, 124. 8. a or a in fat . . . -j Latin, 58. ^ ( English, 128. r Greek, 156. 9. e or e in fret . . . ■< Latin, 129. ( English, 71. f Greek, 75. 10. i ov On fit ....-< Latin, 159. ( English, 241. {Greek, 78. Latin, 38. English, — . {Greek, 11. Latin, 49. English, 58. ( Greek, — . 13. u in fun -\ Latin, — . ( English, 116. III. Diphthongs. 14. au, as in town pro^ iv, f Greek, 13. nounced in the ■< Latin, 3. North ( English, — . {Greek, 9. Latin, — . English, 1. TO THE ARTICULATION OF SYLLABLES. 10] 16. ei or i, as in mine, ( Greek, 46. made up of 13 and s Latin, — . 10 ( English, 75. 17. ai, as in thine pro- ( Greek, 52. nounced in the -j Latin, S3. North ( English, — ( Greek, 33. 18. oi, as in foil . . . •< Latin, — . I English, 5. f Greek, 55. 19. ou, as in cow . . . -| Latin, — . English, 10. 1 The whole sum in each tongue is taken at 1000. 78. Hence we deduce these conclusions : — ■ Remarking that as the Greek is Attic, in Homer the short syllable will be much more frequent, perhaps as 900 : 700 in Virgil. The number of long syllables being taken at 1000, then the number of short will be — in Greek 800, in Latin 762, in English 2303 : which shows at first sight the vast inferiority of our language to the classic, inundated as it is with such an immense disproportion of short syllables. This is much owing, not toAthe native fount of our lan- guage, but to the introduction of Latin words ; such as impossible, opposition, and the like. 79. "We may divide the above list of sounds into three classes : F 3 102 ON THE EXPRESSION DUE (1) The full and hard, comprehending Nos. 1, 2. 5. 14. 17, 18, 19. (2) The full and sweet, comprehending Nos. 3, 4 6. 15, 16. (3) The close and slender, comprehending the short vowels. Then the recurrences of these in the above languages will be, — In Class (1) in Greek, 309, in Latin, 228, in English, 103 (2) „ 249 „ 340 „ 209 (3) „ 444 „ 433 „ 698 And thus we see at once the superior sonorous- ness of the Greek, the os rotundum, and the indis- tinct thinness of the English ; among the sounds of which also, as peculiar to it, comes that num- bered 13, which is more a grunt than a vowel. Its great prevalence is owing to its very common substitution for almost all short vowels in the un- accented syllables ; as, for instance, in temptatwm, and all terminations in ion. A careless pronun- ciation will introduce it more frequently than necessary, and a vulgar will seek it, as in trinutty, anjuls, baptizum, &ic. Nothing shows so much a well-educated ear as avoiding this sound as much as possible without pedantry. 80. The intermixture of these various classes of sounds, by adaptation to the sense, by repetition TO THE ARTICULATION OF SYLLABLES. ]03 immediate, or at intervals, supplies that detail of ornament which fills up the spaces left between the pauses (75). Of course the Greek and Latin poets have made great use of them, as also have the Italian, whose language is so favourable to them. Compare Virgil's Gramina. Nonne vides, croceos ut Tmolus odores. Monstrum horrendum, informe, mgens, cui lumen ademptum, and Petrarch's i^uando io movo i sospiri a chiamar voi E I'onda, che Cariddi assorbe e mesce. See the strong effect of the consonantal syllables in the latter of each of these. The effect is of course greatest when the sylla- ble is marked by the close neighbourhood of a caesura. Thus in that line Fortunam Priami cantabo | et nobile bellum, the opening of the mouth for the broad a just before a principal pause, gives the tantus hiatus ridiculed by the poet. Again, see the quietness produced by the thin and close sounds in Die milii, Musa, yinun captee post tempore Trojse, especially at the csesura. r 4 104 ON THE EXPRESSION DUE Ancient poetry is so rich in well-known examples of this application, that it would be superfluous to quote them. 81. Deficient as the English is in richness and variety of sound, such as it has is set forth in the stronger contrast, and the others have more diffi- culty in finding the close and rugged than we the full and soft. It must indeed be confessed, that if we drew from all the compass of our language, we could never express the two or three short blasts followed by the full outpouring burst of the trumpet, which is suggested in such liveliness by that line — Turn tuba terribilem sonitum procul sere canoro ; and that ineffective indeed is the struggle of Fairfax, the only master of our language among all the translators of Tasso, to convey the imitative expression of his original in the celebrated lines, Jerus. Lib. iv. 3 : — Chiama I'abitator dell' ombre eterne II rauco suon della Tartarea tromba. Treman le spaziose atre caverne, E I'aer deco a quel rumor rimbomba. The dreary trumpet blew a dreadful blast, And rumbled thro' the lands and kingdoms under : Through wasteness wide it roar'd, and hollows vast, And fill'd the deep with horror, fear, and wonder. TO THE ARTICULATION OF SYLLABLES. 105 Although he has contrived, in some degree, tO express the distinction between the sharp blast conveyed by the sound a, and the hollow echo from the deep, signified by the sounds o and u, and their accompanying consonants. Still none can be insensible to MUton's skill in contrasting the proper sounds in the following descriptions of the opening of the gates of heaven and hell, from the sixth and second books, Heav'n opened wide Her ever-during gates, harmonious sound, On golden hinges moving. On a sudden open fly With impetuous recoil, and jarring sound, The infernal doors, and on their hinges grate Harsh thunder. Where the preference of ever-dmring to everlasting seems owing not only to more poetical form, but sweeter sound. The contrast of the vocal sounds in the first, with the consonantal in the second, is very effective. The same poet supplies innumerable examples, as again in Nor on the shaggy top of Mona high. Nor yet where Deva winds her wizard stream. And contrast the pomp of Looks toward Namancos, and Bayona's hold, F 5 106 ON THE EXPRESSION, &C. with the simplicity of Bring the rathe primrose that forsaken lies ; and Pour'd forth profiise on hill, and dale, and plain, with The vassals of his anger, when the scourge ; and Hung high with diamond flaming and with gold, with Grate on their scrannel pipes of wretched straw. Gray, who was among the revivers of attention to our old school, comes next perhaps to Milton in the skilful arrangement of his sounds. The art had become quite forgotten by his time, if we may judge from the gross ignorance shown by Johnson in his criticism on Pope's famous imitative lines. 82. So much for the irregular, though designed mixture of syllabic sounds. But they occur also at regulated intervals, the same articulation or sound being repeated, and forming thus a se- condary ornament in detail, answering to the pri- mary and necessary parts of the structure of verse. 107 CHAPTER XII. ON ALLITERATION. 83. Alliteration means the agreement of syl- lables in their component letters. According to this definition of it, Rhyme is included in it. It is however commonly restricted to the agreement of the consonants of the syllable, and more com- monly still to the agreement of merely the initial consonant. But we are obliged to recur to the fuU and original sense when we speak of it, when adopted as an essential part of versification, as it has been by the modern Welsh, for their most ancient poetry has no such base. The language may have led to it by its singular structure, abounding as it does in similar clusters of letters ; and their later bards riveted its dominion through the usual desire of substituting difficulty of me- chanical execution from poverty of design and scan- tiness of thought. The following is an example of F 6 108 ON AlIilTERATION. tlie construction of one of their most popular measures. See greedy wolves agreed for spoil — rfe&y But d&ilj whets their toil, liord thou orrfainest, and wilt foil With terror all their wild tnrmoil. Poverty of sense in this specimen will be readily excused by such, as reflect on the extreme difiSculty of the attempt in the English, which is as con- trary as possible to the genius of the Welsh in this matter, and that even that tongue cannot go on far without some sacrifice of sense to sound. 84. The definition already given of versification excludes it utterly as a basis (2. 6). And if that argument a priori were not sufiicient, this a pos- teriori would be ample. For surely nothing can be more undignified and inconsistent. Where there is no stress, the effect of such repetition of sound is weak and trifling. Where there is stress, then it becomes as much too strong, as it was before weak, and the result is an intolerable mo- notony, and annoying clink. 85. But one that is unfit for regular service may be very advantageously employed in conducting its lower details, by the irregular operation to which, from their accidental nature, they must be subjected. And such has been the service to which poets of all times and tongues have put ON ALLITERATTON. 109 alliteration. The perception of it is of course most lively (1) when it marks the beginning of a word ; (2) the accented syUahle in a word ; (3) the com- mencement of a foot. We have all these three cases in the line Sed neque meDorum Silvse DiTiSsima Tellus. It adds much to the force of expression in those numerous cases where the notion of repetition is involved. Of these we will mark a few for ex- amples. (1) The first is where the repetition is obviously in the thing itself, as in Ennius's imitation of the note of the trumpet : — At Tuba Terribili aonitu TaraTanTara dixit. (2) Haste commonly presents the idea of a quickly repeated action. Hence Virgil writes, Quo Maxima Motu Terra Tremit, Fugere Ferae. (3) Joy, sorrow, contempt, are expressed by the repetition of the interjections belonging to them, as " ha ! ha ! — ah ! ah ! — pooh ! pooh ! " hence the playfulness of the cattle is expressed by Lucretius in the line, Inde ferse Pecudes Persultant Pabula Iseta j no ON ALLITBEATION. the monotonous iteration of nielanclioly by Tasso in II Pietoao Pastor Pianse al suo Pianto ; and the sneering of contempt by Sophocles in Tu^Xos Ta T' itTa Tov Te vovv Td T' o/i/iaT' ti. (4) Peculiar impressiveness is also conveyed by the same means, as we see in the example of so many of our proverbs, as " Neck or nothing/' Hence Virgil gives the solemn warning of Anchises in the line Neu patriffi Validas in Viscera Vertite Vires. That it should serve on such different, and even contrasted occasions, will surprise no one. The tick of a clock and the prolonged ring of a sheep- bell give a like repetition, but with how different an effect on the mind ! And even the same note repeated under different circumstances wiU give birth to very different feelings. How merry is the chirp of the sparrow in the stackyard, how irksome in the centre of a smoky town. Another use to which Virgil is especially fond of applying it, is to show the connexion of the several parts of the line, and brace together to the mind the portions which the pauses have separated to the voice. Thus at the same time the pauses become much more marked and effective. Indeed ON A1IITERATI01J-. Ill the very idea of connexion seems conveyed in the line MBns AgitAt MOlEm et MAgnO sE cOrpOrB MiscEt. How elaborately simple is the opening of the Georgics : — Quid faciat leetas Segetes, quo Sidere terrain Vertere, Msecenas, ulmisque adjungere Vites Conveniat, quEe Cura bourn, qui Cultus habendo Sit Pecori j aPibus quanta ezFerientia Farcis ; Hinc Canere inCipiam. It is very common with Virgil to have at least two alliterating words in a line ; and when he would give the finishing stroke to an elaborate passage, he commonly does it by a line marked peculiarly thus. What magnificent runs (so to say) are closed by the following lines : — RomanOS ad Templa Deum Duxere TriumphOS. — Georg. ii. 148. Et SeptemgemiNI Turbant Trepidi oSTia NIII.— .lEn. vl. 801. So little did this master of expression slight this auxiliary, which has been treated so cavalierly by modern critics. Indeed sometimes it is not easy to assign his reasons for such excessive use of it, as in Et quae MarMOReO fert MONstra sub aquORe pONtus. Ma. vi. 729. CorPorEse ExCEdunt PEStES ; PENitusque NECEsse Est. Ma. vi. 737. 112 ON ALLITERATION. Our poet Gray, who, next to Milton, had the finest ear, has similarly employed it, as in the line " Ruin seize thee. Ruthless King." Johnson, with his usual ignorance, has condemned him for it. It has, without doubt, been abused. But the very abuse proves its use in moderation. Nor should we judge of its general effect from our own lan- guage, which is so tough and harsh from its quan- tity of monosyllables. 113 CHAPTER XIII. ON BHTMB. 86. Rhyme, in its original fonn Rliythm, means proportion, and among other kinds Metrical, in •which sense alone it occurs in the modern tongues which have borrowed the word. But as in these tongues the great distinction of one verse from another is formed by the identity of the ter- minating sound, it came to be applied as the term for this identity, and from familiar use assumed the more familiar form of rhyme. Not that the whole syllables which have the same rhyme are identical. For the initial consonant is left indif- ferent. It would be a work of insuperable diffi- culty, as well as an effect most barbarous, to include that also, and make reclaim answer to declaim. 87. As an auxiliary, it has been adopted in the earliest poetry with which we are acquainted, in that of Homer, and runs through all classical poetry to the last. Its use is to give strength to 114 ON KHTME. csesuras, which are made to rhyme with the end of the line. Most commonly the termination of the substantive rhymes with that of its adjective, whose relation thus becomes pointedly marked. Who is not familiar with such examples as 'E« iJikv K.pr]Tdu)v \ y&voQ €VX0fiai tiipEidiov. — Odyas. JSJ. 199. Ter centum nivei | tondent dumeta juvenci, Nitor, et indicio | prodor ab ipse meo. Ilium indignanti | similem, similemque minanti. 88. Those who censure rhyme as a modem bar- barism do not seem to have considered this use, nor that, it would be exceedingly barbarous in our own versification in that very place where it is such a favourite with the ancients ; for who could endure such a line as The blushing rose | her dewy flower shall close ? for our rhymed verse has rhyme enough already, and the whole couplet would be thrown into con- fusion. And our blank verse rejects every thing of the kind on principle. 89. That line Ilium indignanti similem, similemque minanti, is the very model of the Leonine or monkish verses, and gives us the key to the origin of modern rhyme. For when in the decay of the Latin the sense of quantity had been lost, and the musical ON RHYME. 115 accent given way to mere stress, the hexameter lost all its ancient music, and could well bear some addition to its harmony. At the same time the ear had become dull to the variety of its pauses, and the very marked caesura in the middle of the third foot was the only one to which it was fully sensible, and which therefore it always required. But at this very point also came the ancient rhyme. What could be a more natural resource for patch- ing up the ruined harmony than this, and where could they have found a better if they had pur- posely looked beyond this which was at hand? When therefore the vernacular tongues began to employ versification, how could it be otherwise, decayed Latin as they were, than that they should have recourse to rhyme, feeling as they must the want of some additional prop to their feeble measure. The argument is much strengthened when we find, from specimens of the Anglo-Saxon, that it was unlikely to have been derived from the Teutonic stock. 90. The capability of languages for rhyme is of course very difi'erent ; being in proportion to their analogical structure and abundance of inflections. Hence the Greek surpasses all on this point. It abounds with words of similar sound, as SpvirToi, Opvirrw, KV1TTW, tutttw ; — of similar derivation, as Opin/ia, TTE/ujua, arififia, arpififia ; — with large classes 116 ON RHYME. of various terminations of substantives, as rj/xa, Wfia, rime, wcrig, icrfia, laig, &C., — of adjectives, as r)p?)C, Jjpoc, tpoe, ivog, wSrjQ, riXog, &C. : to say nothing of the inexhaustible resources arising from persons, tenses, and cases. Even the modem Greek, shorn though it be of so much of this richness, is still by far the best furnished of all our European tongues with the means of rhyme. 91. As in every respect, so in this the Latin is inferior to Greek. Being far less homogeneous in its structure, it has not such well-fiUed classes of various inflection, and so much echo resounding from an analogical structure. Still it is richer than all its daughters ; which, if they have laid open a wider field to a given number of rhymes, by re- ducing many inflections to one, have lost variety in proportion. Thus, where the Latin had terror, terroris, terrorem, terrori, terrore ; the Italian has but terrore : and, therefore, with five times as many words to one rhyme (which it does not need), it has but one- fifth of the variety which it does need. The same may be said of the Spanish and of the French, which last is the poorest of the Latin dialects on this point j while, through its monosyllabic inflexions, it departs widely from their scope of variety, which admits of choice between the fulness of the double, and the firm- ness of the single, rhyme. ON EHYME. 117 92. Unquestionably the poorest of all the tongues of modern literature, on this head, is the English : not only from its very compound nature, which is so unfavourahle to analogical structure, and to rich variety of inflexion ; and from its monosyl- labic terminations : but also, and principally, from the natural position of its accent, which can hardly ever fall upon a termination. For example, while the Italian can match orrore with terrore, and the French, vicieuso with melodieux; the English cannot put horror to rhyme with terror, nor vicious with melodious ; and these terminations, which we have so largely introduced from the French, and are to that tongue so fruitful a resource for rhyme, are to us quite barren. Not, however, that they were such at the time of their introduction ; for they seem to have brought with them the native seat of their stress. Thus, in the opening of the " Canterbury Tales," we find among the rhymes of the first two hundred lines, liquor, reason, gippon, voyctge, visage, hracer, manere, conscience, Bennet, labour. But they were in time compelled to conform to the genius of their adopting language, which likes to throw the stress upon the radical part of the word, and is especially averse to its place on the last syllable. Hence our rhymes are driven to the scanty resource of primitives and radicals : and while the Italian and Frenchman, having lis ON EHYME. ended his verse with an infinitive or participle, has the choice of a thousand more such for a rhyme to it, the Englishman must often go through his alphabet, and often congratulate himself, that, amid the identical sounds thus studiously sug- gested, there does occur one which is contained in a word suitable to the sense required. The extreme difficulty of the case wiU appear from another consideration. We have seen, that in the whole compass of the language, the recurrence of the long syllable is to that of the short but as 10 : 23 (78) ; but in rhymes, it is as 84 : 10, or nearly 70 : 23 : to such a degree must the nature of the language be forced to supply them. 93. But hence also arises a great advantage to the English. For it is reasonable that the close of a verse should not be the least emphatic part of it, but just the contrary. But this defect of em- phasis is the case with both Greek and Latin, and their children where rhyme is used, since that commonly falls upon mere inflexions. No wonder that it should produce a mere tinkle in ears accustomed to classical versification. But with us, where it is single (as it almost invariably is, except when a ludicrous or trifling efiect is intended) it can never fall on a syllable more weak and insignificant than the rest. Hence, with us rhyme assumes a much more dignified ON RHYME. 119 position, and derives additional strength and beauty from the very difficulty of its production. And this cause, added to the strictness of our rhythm, allows us to he satisfied with it even when imperfect, and within certain bounds to enjoy the imperfection for its agreeable variety. In the two most perfect specimens of our rhymed couplet, Pope's "Eape of the Lock," and Par- nell's "Hermit," one rhyme in twelve is imper- fect. In a significant thing we can endure inex- actness, for we are not always inclined or able to appreciate the fulness of its significance. But an insignificant thing is comprehended at first sight, and must be perfect to be endurable. Voltaire well remarked that rhyme was a master to the French, but a slave to the English. 94. Thus rhyme in English goes on quite a different footing from that which it mainly holds in the Latin dialects. A word is echoed by a word, and not a termination by a termination. In the former case, the same sound conveys an- other sense, and thus brings also a pleasing sense of variety, often of analogy, to the mind, which finds itself transported by it to another region of thought. In the latter, there is a mere repetition of senseless sound, and the mind finds itself simply in the same part of the accidence. An example will make these remarks clear : — 120 ON EHYME. II est certains esprits, dont las sombres pensees, Sont d'un image epais toujours embarassees ; Le jour de la raison ne la sauroit percer. Avant done que d'ecrire apprenez a penser. Selon que notre idee est plus ou moins obscure L'expression la suit, ou moins nette, ou plus pure. Ce que Von con^oit bien s'enonce clairement, Et les mots pour le dire arrivent aisement. Sur tout qu'en vos ecrits la langue reveree Dans vos plus grands exces vous soit toujours sacree. En vain vous me frappez d'un son melodieux, Si le terme est impropre, ou le tour vicieux. Boileau, I'Art Poetique, i. Cosi pugnato fu, sinche I'albore Rosseggiando nel del gia n'apparia. Ma poi cbe scosso fu il uotturno orrore, Che Torror deUe morti in se copria, La desiata luce a noi terrore Con vista accrebbe dolorosa e ria ; Che pien d'estinti il campo, et quasi tutta Nostra gente vedemmo ormai distrutta. Gerusalemme Lib., viii. 20. Now sank the sun. The closing hour of day Came onward, mantled o'er with sober grey. Nature in silence bade the world repose ; When near the road a stately palace rose. There by the moon through ranks of trees they pass, Whose verdure crown'd their sloping sides of grass. It chanced the noble master of the dome Still made his house the wandering stranger's home ; Yet still the kindness, from a thirst of praise, Proved the vain flourish of expensive ease. The pair arrive : the liveried servants wait : Their lord receives them at the pompous gate. The table groans with costly piles of food. And all is more than hospitably good. ON RHYME. 121 Then led to rest the day's long toil they drown, Deep sunk in sleep, and silk, and heaps of down. ParnelVs Hermit , 44. 95. So very strongly marked as it is, rhyme becomes a mere tinkling, and loses all dignity either if the return occur too soon, or if it be marked with a distinctness too elaborate. The short verses of Skelton, such as, Thus I, Colin Clout, As I go about, And wondering as I walk, I hear the people talk, are an example of the first case. Nor is sufficient solemnity for sustaining serious narrative obtained until we come to the heroic couplet. And even here the eifect is not majestic enough for the march of epic poetry, as every one must feel on reading the translations of Dryden and Pope. Spenser, therefore, did well in introducing from the Italian the alternations of the stanzaic construction. As to the second case of too studied a rhyme, its light and commonly ludicrous effect appears in the triple endings of the Italian, and in the double endings of our lines, as in, Who now to sense, and now to nonsense leaning. Means not, but blunders round about a meaning. Perhaps a better specimen could not be found than these lines of Fairfax : — 122 ON RHYME While thus he spake, Erminia, huah'd and still, His wise discourses heard with great attention ; His speeches grave those idle fancies kill Which in her troubled mind had such dissension. After much thought reformed was her will ; Within those woods to dwell was her intention, — The Recovery of Jerusalem, vii. 14. where the reader will not fail to remark also the very bad effect of the rhyme falling on the same termination in our tongue. It does not seem to us a rhyme, hut a mere repetition. 123 CHAPTER XIV. ON THE MODERN EECITATION OF ANCIENT VEESE. 96. That we enjoy a perception of great harmony when we recite the ancient measures is indis- putable ; as also that such harmony very much sui-passes any that we can derive from all the stores of modern versification. And yet we must derive it from a source quite other than the ori- ginal and true, which was the proportion of quan- tity. For let any one try to recite (if he can) a line, strictly according to its long and short syllables : he will find in it rather the reverse of what he has been accustomed to consider as har- mony, and at least his attention will be diverted from the pauses. Let him again make the near- est approach to quantity that stress can give, by laying it on all the long syllables. This has no foundation in the position of accent or quantity, and the result is nothing but the intolerable jingle, which reminds him of his days of scanning schoolboyhood. And here, again, he has no pro- G 2 124 ON THE MODERN RECITATION per perception of the pauses. Let him, lastly, substitute stress for the Latin accent, and recite accordingly in the usual way. He finds a most agreeable harmony, and is delighted with the rich variety of the pauses. And all is seemingly as unaccountable as agreeable ; for all the rules of quantity have been quite overturned, and even directly contradicted. Iambuses and spondees are turned into trochees, as in novis, causam; tribrachs, anapaests, and cretics into dactyls, as in fdciat, s^getes, Candidas. And two dissyllabic words in sequence, as T^mpla Beum, Bev/m Templa, saevw feres, dlta juga, &c., compose what feet they will according to quantity, are to us all double trochees ; and two trisyllabic will either form the double dactyl, as in s^getes apibus, mur- murant latices ; or a combination of the amphi- brach and dactyl, as in hahendo pecori. Hence it comes to pass that our modern recitation con- veys no analogous notion of any ancient Latin measure, except of the trochaic quite pure, of the dactylic quite pure, and of the latter only in those cases where the foot is contained in a word, or in the combination of the trochee and amphibrach, represented in cuUus hahendo. But one real dac- tyl can stress substituted for accent bring out of the wholly dactylic hexameter, Quadrupedante pedum sonitu quatit ungula campum. OF ANCIENT VERSE. 125 Even as to our sense of the iambic measures, which are most congenial of all with our notions of harmony, forming as they do the staple of our own versification, we are continually "at fault owing to the nature of the Latin accent, which will not allow a dissyllable to be an iambus, but makes it a trochee. Hence a multitude of lines are to us scazons, and we even go so far as to end a trimeter with three trochees ; thus in Horace, Epod. xvii. 37; xviii. 17. Effare, jussas cum i^de poenas luam. In monte saxum. Sed vetant leges Jovis. If we open Horace at the first Ode, we shall find that we make the trochaic movement in " Nauta secet mare," equivalent to the dactylic in " fidite rdgibus ;" again, in " Laudat rura sfii," to the iambic in " Myrtdum pavidus ; also, the iambic movement " Maecenas atavis," to be responded to by the trochaic " Evitata rdtis :" and in general the dactyl in quantity may be in accent an am- phibrach, as in conveniat, or a cretic, as in fudit equum; and thus suit the iambic or trochaic measures instead of the dactylic. However, therefore, we must learn to compose the ancient measures according to the ancient prosody, yet we must have recourse to a very different principle in considering the rules of the g3 126 ON THE MODERN RECITATION modern recitation of them. This seems hardly to have been discussed, and assuredly never has been ascertained. 97. Of course, the basis of it, which is the stress laid on the seats of the Latin accent, must be one which lies at the foundation of the pre- vailing rhythm of our modern tongues. And this as we have already seen (15), is the iambic measure. For example, according to our pronun- ciation, the Sapphic, Arte materna rapidos morantes is a good dramatic line, running as. Questions important agitate debaters with the common licence of the trochee in the first place. But it will also supply in its first three feet and half as many to the commence- ment of an accentuated hexameter, as in, Arte matema rapidos | per saxa morautem ; or of a trimeter, such as in Hor. Epod. vii. 5, N6n ut superbas invidse | Carthaginis. And the very dactylic cast of the line, Conveniat quae cura bdum qui | cultus babendo, can be represented exactly in its first four feet, by our heroic, as in. Convenience, which ever orders, is | always directing. OF ANCIENT VEESE. 127 And, again, that of Liber et alma Ceres, vestro si | numine tellus, by Bacchus and joyous Ceres under your | Offices kindly. In many cases also tte trimeter, by the simple addition of the one syllable which is wanting to make it equal in the number of syllables to the hexameter, which has but the one dactyl, be- comes an hexameter. Of the thirty-three tri- meters which enter into Horace's sixteenth Epode, seven thus become hexameters, namely, verses 2, 4, 10, 20, 48, 52, 60. Thus, Neque impudica Colchis (hue) intulit pedem, is as to accent as perfect an hexameter as Namque indef6s90s artus non dpprimit aetas. 98. The iambic then seems to be at the foun- dation of all the harmony which we can make out of the chief ancient measures according to our recitation, except in the case, of course, of the trochaic ; which, however, as we have seen, we mix most strangely with the iambic, so as to go quite contrary to the rule of quantity, which sets these two measures in as opposite directions as possible. This mixture it will be necessary to g4 128 ON THE MODERN RECITATION bear in mind continually, in laying down the rules of recitation for certain measures. How we not only tolerate it, but indulge in it, were unac- countable, if we did not consider the peculiar nature of stress, indefinite as it is in comparison with accent or quantity. It is in fact merely accidental to them. We may pronounce one syl- lable in twice the time that we do another, or two or three higher in the scale than we do another, either in a louder or a softer tone, and either as to all or to any particular syllables, and yet not essentially disturb the proper proportions, though it may require some slight care not so to do. And while such proportions are so fixed, stress admits of none. One sound is louder to us than another, but we never can distinctly perceive it to be so in any proportion. We cannot fix our loudest notes at a given height above our com- mon tone^ Hence, as we have seen formerly (1 6), there may be numberless degrees of stress, and we may overcome a lower stress by a higher one. Thus, in beginning an hexameter with "f6ns erat," we do not neglect the stress on erat, but we ignore it (so to say) by a more forcible one on fans. Thus we can account for such lines as By fowl, fish, beast, was flown, was swnm, was walk'd. Par. Lost, vii. OF ANCIENT VERSE. 3 29 Beast, bird, fish, insect, what no eye can see. Essay on Man, i. Amid such indefiniteness it is most easy to lose our reckoning, and take our start as it -were from the louder tap, and so turn the iambic movement into trochaic, or to end with the loud tap, and turn the trochaic into iambic. However it be as to cause, the effect is certain, as we have seen in the examples above given. 99. It must never be forgotten that our recita- tation of the two ancient languages goes entirely by the rules of the Latin accent. Those of the Greek are utterly unattended to. Thus we pronounce vnipa, OvriTOQ, iTri(3i\iiroiv, as if they had been accentuated as rifiepa, dvriTog, iTri^Xiiroov, accord- ing to the Latin rule. Hence we have not the slightest sense of Greek recitation in any waj whatever, and in pronouncing Homer's verse as we do, we translate him into our pronunciation of Virgil's. Indeed, it were impossible by any sub- stitution of stress in place of the Greek accent, to arrive at the slightest analogy of the construction of the hexameter, of which we think that we have some sense according to the Latin, inasmuch as the concluding dactyl and spondee can be adequately represented. For the Greek accent, not depending on the quantity of the penulti- mate, but rather of the ultimate, will not present G 5 130 ON THE MODERN RECITATION, &C. the analogies of the dactyl and trochee, and also frequently falls on the last syllable. How shock- ingly we violate its rules may be seen from the opening of the Iliad, in which we read perforce, dsiSt ov\6fi£vtiv, I0))ic£, kuve'