MISCELLANEA THE LONG UM AND BICEPS OF THE GREEK HEXAMETER Recent articles by Irigoin 1) and Devine & Stephens 2) have demonstrated that the statement of Dionysius of Halicarnassus that the monosyllabic biceps of the Greek hexameter is 'longer' 3) than the longum is basically true 4). Irigoin concluded that the phonetic structure of the longunz was different from that of the biceps. Devine & Stephens deny this, and attribute the greater length of the biceps to the effect of the metrical structure of the hexameter on the Greek language. There is, of course, a basic difference between the longum and the biceps : the former cannot be resolved, the latter can. Devine & Stephens are right, however, to suggest that the structure of the hexameter is usually responsible for a shorter longum. This becomes very clear when we consider such factors as lengthening in arsis, hiatus in arsis, short final vowel lengthened in arsis before mute and liquid, and so on. These features are mentioned by West 5) as evi- dence that the thesis (biceps) is longer than the arsis (longum). If the arsis were as long as the thesis, according to West, there would be no reason why such factors should not occur as readily in thesis; they are tolerated in arsis because the arsis is shorter. This state- ment, however, overlooks a fundamental factor in the structure of the Greek hexameter: the restrictions on word-end in thesis. The factors mentioned by West generally involve word-end, the length- ened syllable usually being final. The possibility of 'lengthening in thesis' obviously presupposes a monosyllabic thesis; but word-end after a monosyllabic biceps is greatly restricted in the hexameter 6). Consequently there is actually very little opportunity for the factors mentioned by West to occur in thesis. A related point is that a dactyl is at least twice, often more, as frequent as a spondee in Greek hexameter authors, hence the relative infrequency of mono- syllabic bicipitia. By contrast there are few restrictions on word-end in arsis. It should be stressed that the restrictions on word-end after monosyllabic thesis do not result from the fact that the thesis is longer. They have nothing to do with it. Word-end, in fact, is allowed far more frequently after a disyllabic biceps, which is cer- tainly not shorter, and is possibly longer, than a monosyllabic bicefis. 69 So the fact that the biceps is longer than the longum seems irrelevant to word-end. While, then, the longer biceps is connected with the metrical structure of the line, it is unwise to deny that the poets themselves also contributed to the situation. Let us consider, for instance, Theocritus' use of the infinitive endings -gw and -ev in his Doric poems. Where the mss of Theocritus give both forms, and where both are metrically possible-i.e. when the following word begins with a consonant-Gow 7) tells us that there is no way of choosing between them. He writes: "The nuances which might decide be- tween ('Xe'L3e&vand ... are now too subtle to be grasped, and if we were sure, as we are not, which Theocritus chose, we could not appreciate the reasons for, or the effect of, his choice". But we see that -eLv is than -ev and since we know that the thesis can longer is longer than the arsis, we might expect, all other things being equal, to find -ev in arsis, -ELVin thesis. Now the mss of Theocritus frequently differ among themselves, and often offer both forms, but, on Gow's own admission, where the metre is not affected, "the mss are apparently unanimous for -ev in six places". It should come as no surprise to learn that in all six places -ev is in arsis. This must surely represent what Theocritus wrote, guided by his feeling for the metrical arrangement of the hexameter; this, of course, need not have been the only reason, but it was surely a contributing factor. A further feature worth considering is paragogic nu. It is clear from Isler 8) that the use of paragogic nu to lengthen a short syllable is again a matter of choice rather than due to the restrictions of the hexameter in conjunction with the Greek language. This becomes evident from the differences between the poets in their treatment of paragogic nu. Homer, for instance, uses it frequently to lengthen the arsis (1071 examples), but not so frequently the thesis (208 examples): from then on there is a gradual decline in its use for lengthening purposes in both parts of the foot. Callimachus, in his Hymns, has twenty-two examples in arsis, none in thesis; Nonnus, in nine books of his Dionysiaca, has only one example in arsis, none in thesis. So there is an increasing tendency to avoid using para- gogic nu for lengthening in thesis (never in Callimachus, Oppian, Quintus Smyrn.) and in arsis, unlike the optimal lengthenings involv- ing mute and liquid etc., mentioned above. It seems that there was a feeling that paragogic nu was not regarded as a fully authentic consonant. Its optional nature must have made poets feel that its proper purpose was to avoid hiatus, and not to make position. Isler's evidence certainly shows that there was no similar decline in the use of paragogic nu to avoid hiatus from Homer to Nonnus 9). .