Lines, however, where words are left entire, without being divided even by a femipaufe, run by that means much the more fweetly:

Quadrupedante putrem || fonitu quatit | ungula campum. Again;

Eurydicen toto referebant | flumine ripæ,

The reafon of thefe obfervations will be evident up. on the flighteft reflection. Between things fo intimately connected in reading aloud, as are fenfe and found, every degree of difcord is unpleasant: and for that reafon, it is a matter of importance, to make the musical paufes coincide as much as poffible with thofe of fenfe; which is requifite, more efpecially, with refpect to the paufe, a deviation from the rule being lefs remarkable in a femipause. Confidering the matter as to melody folely, it is indifferent whether the paufes be at the end of words or in the middle; but when we carry the fenfe along, it is difagreeable to find a word split into two by a paufe, as if there were really two words: and though the difagreeablenefs here be connected with the fenfe only, it is by an eafy tranfition, of perceptions transferred to the found;

found; by which means we conceive a line to be harfh and grating to the ear, when in reality it is only fo to the understanding.*

To the rule that fixes the paufe after the fifth portion, there is one exception, and no more: if the fyllable fucceeding the 5th portion be fhort, the paufe is fometimes poftponed to it.

Pupillis quos dura pre mit cuftodia matrum,

Again,

In terres oppreffa || gravi sub religione.

Again:

Et quorum pars magna || fui; quis talia fando,

This contributes to diverfify the melody; and where the words are smooth and liquid, is not ungraceful; as in the following examples:

Formofam refonare | doces Amaryllida fylvas.

Again:

Agricolas, quibus ipfa procul difcordibus armis.

If this pause, placed as aforefaid after the fhort fyllable, happen alfo to divide a word, the melody by thefe circumftances is totally annihilated. Witnefsthe following line of Ennius which is plain profe.

Romæ monia terruit impiger | Hannibal armis.

Hitherto the arrangement of the long and fhort fyllables of an Hexameter line and its different pauses, have

*See chap. 2. part 1. fe&t. 5.

have been confidered with refpect to melody, but to have a just notion of Hexameter verfe, these par ticulars must also be confidered with respect to fenfe. There is not perhaps in any other fort of verfe, fuch latitude in the long and fhort fyllables; a circum stance that contributes greatly to that richness of melody which is remarkable in Hexameter verse, and which made Ariftotle pronounce, that an epic poem in any other verfe would not fucceed.* One defect, however, muft not be diffembled, that the fame means which contribute to the richness of the melody, render it lefs fit than feveral other forts for a narrative poem. There cannot be a more artful contrivance, as above obferved, than to close an Hexameter line with two long fyllables preceded by two fhort but unhappily this conftruction proves a great embarrassment to the fenfe; which will thus be evident. As in general, there ought to be a strict concordance between a thought and the words in which it is dreffed; fo in particular, every close in the fense ought to be accompanied with a close in the found. In profe this law may be ftrictly obferved; but in verfe, the fame ftrictnefs would occafion infuperable difficulties. Willing to facrifice to the melody of verfe, fome fhare of the concordance between thought and expreffion, we freely excufe the feparation of the mufical paufe from that of the fenfe, during the courfe of a line; but the clofe of an Hexameter line is too confpicuous to admit this liberty: for which reafon there ought always to be fome pause in the sense at the end of every Hexameter line, were it but fuch a pause as is marked with a comma; and for the fame reafon, there ought never to be a full clofe in the fenfe but at the end of a line, becaufe there the melody.

* Poet. cap., 23,

melody is clofed. An Hexameter line, to preferve its melody, cannot well admit any greater relaxation; and yet in a narrative poem, it is extremely difficult to adhere ftrictly to the rule even with thefe indulgences. Virgil, the chief of poets for verfification, is forced often to end a line without any clofe in the fenfe, and as often to close the fenfe during the running of a line; though a close in the melody during the movement of the thought, or a clofe in the thought during the movement of the melody, cannot be agreeable.

The accent to which we proceed, is no less effential than the other circumstances above handled. By a good ear it will be difcerned, that in every line there is one fyllable diftinguishable from the reft by a capital accent: that fyllable, being the 7th por tion, is invariably long.

Nec bene promeritis | capitûr nec | tangitur ira, Again :

Non fibi fed toto genitûm fe credere mundo. Again:

Qualis fpelunca fubitô commota columba.

In thefe examples, the accent is laid upon the laft fyllable of a word; which is favourable to the melody in the following refpect, that the paufe, which for the fake of reading diftin&tiy muft follow every word, gives opportunity to prolong the accent. And for that reafon, a line thus accented, has a more fpirited air, than when the accent is placed on any other

other fyllable. Compare the foregoing lines with the following:

Alba neque Affyrio || fucâtur | laná veneno.

Ágain :

Panditur interea domus ômnipotentis Olympi.

Again :

Olli fedato | refpondit. corde Latinus.

In lines where the paufe comes after the fhort fyl lable fucceeding the fifth portion, the accent is dif placed, and rendered lefs fenfible: it feems to be fplit into two, and to be laid partly on the 5th portion, and partly on the 7th, its ufual place; as in

Nuda genu, nodôque | finûs collecta fluentes. Again :

Formofam refonâre | docês Amaryllida fylvas.

Befide this capital accent, flighter accents are laid upon other portions: particularly upon the 4th, unlefs where it confifts of two fhort fyllables; upon the 9th, which is always a long fyllable; and upon the 11th, where the line concludes with a monofyllable. Such conclufion by the by, impairs the melody, and for that reafon is not to be indulged unlefs where it is expreffive of the fenfe. The following lines are marked with all the accents.

Ludere quæ vêllem calamo permifit agrefti.