for its own fake. Accordingly, we find, that, though we cannot fo eafily trace that accentual rhythmus which forms the harmony of the beginning and middle of a sentence, yet the latter part, or what is commonly called the cadence, confifts (when harmoniously conftructed) of fuch an arrangement of accented words as approaches nearly to verfe. Every ear will immediately find a ruggedness and want of harmony in the conclufion of the following fen tence: We are always complaining our days are few, and acting as though there would be no end of them. Addison. The reason of this harshness seems to be, that vaft chaẩm of unaccented words that extends from the word acting to the word end. The ear, indeed, fenfible of the want of accent, lays a little ftrefs upon though: but this does not quite remedy the evil; ftill there are four words unaccented, and the fentence remains harsh: but if we alter its ftructure, by placing a word that admits of an accent in the middle of these four words, we fhall find harmony fucceed to harshness and inequality. We are always complaining our days are few, and acting as though there would never be an end of them. This difference, therefore, can arise from nothing but an unequal and unmetrical arrangement of accent in the former fentence, and a greater approach to equal and metrical arrangement of accent in the latter. As a farther corroboration of the truth of this opinion, let us take a fentence remarkable for its harmony, and try whether it arifes from the foregoing principles. We hear at this diftance but a faint echo of that thunder in Demofthenes, which shook the throne of Macedon to its foundations; and are fometimes at a lofs for that conviction in the arguments of Cicero, which balanced in the midft of convulfions the tottering republic of Rome. In the latter part of this fentence, we find the accented fyllables at exactly equal intervals from the word fometimes to the word midft; that is, there are three unaccented fyllables between every accented fyllable: and from the word midft to the word Rome, there is an exact equality of intervals; that is, two unaccented fyllables, or, which is perfectly equivalent, fyllables pronounced in the time of two, to one unaccented. Now, if we change a few of the words of this fentence to others of different length and accent, we fhall find the harmony of the fentence confiderably diminished, though the sense may be inviolably preserved. We hear at this distance but a faint echo of that thunder in Demofthenes which fhook the throne of Macedon to its foundations; and are sometimes at a loss for that force in the proofs of Cicero, which balanced in the midst of anarchy the tottering ftate of Rome. That full flow of profaic harmony, fo perceptible in the former fentence, is greatly diminished in this; and the reafon feems plainly pointed out for as the harmony of verse is owing folely to an equal and regular return of accent, the harmony of profe muft arife from the fame fource; that is, as verfe owes its harmony entirely to a regular return of accent, profe can never be harmonious by a total want of it. The fole difference between them feems to lie in the conftant, regular, and artificial arrangement of accent in the one, and the unftudied, various, and even oppofite arrangement in the other. Verfe, with fome few exceptions, proceeds in a regular alternation of accent, from one end of the poem to the other; harmonious profe, on the contrary, in fome members, adopts one fpecies of arrangement, and in fome another; but always fo as to avoid fuch clusters of accents in one place, and fuch a total abfence of them in another, as neceffarily occafions a ruggednefs and difficulty of pro nunciation. At first fight, perhaps, we fhould be led to fuppofe, that the intervals between the accents ought rather to diminish than increase as they approach the end of a fentence; and yet, if we confult the ear, we fhall find that intervals. of two unaccented fyllables found better even in the clofing member of a sentence, than intervals of one unaccented fyllable only. Let us take the following fentence as an example of this: Demetrius compares profperity to the indulgence of a fond mother to a child, which often proves his ruin; but the affection of the Divine Being to that of a wife father, who would have his fons exercifed in labour, difappointment, and pain, that they may gather ftrength and fortitude. Now, if, instead of the word ftrength, we subftitute experience, though the fenfe may be weakened, the found will, perhaps, be improved; and if the ears of others fhould agree with mine in this particular, it may be laid down as a rule, that other circumftances being equal, the laft members of fentences ought rather to end in the dactylic than in the iambic measure. In this appellation of the measure of profe, I adopt the terms generally made ufe of, and particularly by Mafon, in his Effay on Profaic Numbers. This gentleman deferves much praise for his attempt to inveftigate the caufes of profaic harmony, but appears to me to have an idea of English metre fo blended with that of the Latin and Greek, as to throw error and confufion over his whole performFor what can we make of his placing two long quantities over the two fyllables of the words sentence and spōndee? Each of thefe words can have but one accent; and it is accent, or emphafis, and these only, and not any length or openness of the vowels, that forms English metre, or that rhythmus which is analogous to it in profe. ance. Harmony of Profaic Inflexions. Hitherto I have only confidered poetic and profaic harmony as arifing from an harmonious and rhythmical arrangement of accent; and it is with fome diffidence I venture upon a farther explication of this fubject upon principles which have never yet been thought of: but I prefume it will be found, upon inquiry, that the various and harmonious arrangement of the rifing and falling inflexions of the voice, is no lefs the cause of harmony, both in verse and profe, than the metrical arrangement of accent and emphasis. The melody both of profe and verse seems to confift as much in fuch an arrangement of emphatic inflexion, as fuits the fenfe, and is agreeable to the ear, as it does in a rhythmical difpofition of accented and emphatic fyllables. To illuftrate this obfervation, let us take an harmonious couplet in Pope's Prologue to Cato: A bráve man ftruggling in the storms of fate, The first line of this couplet ends with the rifing inflexion, to prevent the want of harmony there would be in ending two fucceffive lines with the fame inflexion; a fameness for which nothing but emphafis will ever apologize. As this line ends with the rifing inflexion, the last word may not improperly be called the rudder, which directs the inflexions on the preceding words; for in order to prevent an exact return of the fame order of inflexion, it is not sufficient that the different inflexions fucceed each other alternately; this would be like the fucceffive founds of the letters A, B; A, B. To prevent a return of founds fo little various, we find the ear generally adopt a fucceffion of inflexion, which interpofes two fimilar inflexions between two fimilar inflexions; and this produces a variety fimilar to the series, A, B, B, A; or B, A, A, B. The firft line, therefore, of this verse, neceffarily ending with the rifing inflexion on the word fate, in order to make the other words as various and harmonious as poffible, the falling inflexion is placed on forms, the fame inflexion on ftruggling, and the rifing inflexion on brave; and this, in the firft line, forms the arrangement, rifing, falling, falling, rifing; or A, B, B, A. The next line ending the fentence, neceffa |