STOP Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org. THE IAMBIC TRIMETER IN MENANDER By John Williams White Pure trimeters, that consist solely of iambs, are comparatively rare in Menander. Only 18 of the 728 trimeters now taken into account 1 are pure (1 in 40), but the number is relatively larger than in Aristophanes (1 in 69 ). 2 The plays of Aristophanes, however, vary in this particular: the Nubes contains the greatest number of pure iambic lines, 18 in 758 non-lyrical trimeters (1 in 42), the Equites the fewest, 6 in 688 (1 in 115). 3 Irrational and trisyllabic feet, in place of the normal iamb, are extremely common in both poets, and verses occur in which no 1 The tables and other numerical statements relating to the trimeters of Menander that are included in this treatise are based upon those trimeters in the recently dis- covered Cairo MS which have metrical form that is assured, although they are not all complete in the MS. The four plays are designated as I ("Hpws), II ('E7rii7>^7rojTes), III (HepiKeipoi*4rr}), IV (2a/i(a). The numbering of verses is that of the princeps. The trimeters under consideration are the following : I. 20-26, 36-56, 58-61 ; II. 1-9, 11-21, 23-51, 54-84, 86-91, 93-97, 99-103, 105, 107, 109-17, 119-20, 123-28, 130-38, 141-49, 151-52, 159-72, 183-84, 186-87, 189, 191, 194-210, 212-14, 216-22, 224, 227-38, 240-43, 247-48, 250-57, 259, 262, 264-78, 280, 283-92, 294-96, 298-300, 302-3, 305-16, 318-32, 334-39, 341-49, 351-53, 355-56, 358-59, 370, 372-76, 378-81, 383, 385-86, 392-404, 406-17, 419-23, 446-50, 452-54, 456-59, 461-77, 482-506, 509-13 ; III. 7-22, 24-27, 29-30, 32, 36, 39-43, 45, 47-49, 51-70, 72, 76-80, 84-93, 95-101, 103-7, 110, 113-14, 116-18, 120-22, 124-30, 132-38, 140-41, 146, 165-67; IV. 4-9, 11-13, 16-17, 19-32, 37-51, 53-64, 74-80, 85-88, 90-92, 98, 107-14, 116, 120-22, 132-35, 137-59, 161-63, 168, 170-76, 179-82, 184-91, 193-201, 271-72, 277-87, 290-93, 296-302, 304-6, 313-17, 319, 321, 342-45, 435-39, 442-45, 447, 450, 453, 470-71. These 728 verses furnish the facts for the tables and for all general statements. In determining the laws that govern the use of trisyllabic feet, I have adduced facts from the incomplete lines also, when this was necessary. Occasionally these lines are cited for other purposes. The numbers referring to such trimeters are starred. 2 In Aeschylus 1 in 14, in Sophocles 1 in 17, in Euripides 1 in 22.5. See J. Rumpel " Der Trimeter des Aristophanes," in Philology* XXVIII (1869), p. 601. For the poets of the Middle and New Comedy, see P. Perschinka, " De mediae et novae quae vocatur comoediae Atticae trimetro iambico," in Dissertationes philologae Vindobonenses III (1881), p. 329. Citations of evidence from the fragments of Menander in the present article are drawn from Perschinka's dissertation, except those that relate to anapaests. The facts about anapaests have been independently collected. 3 The numerical statements and tables relating to the trimeter of Aristophanes that are included in this treatise, in order to furnish means of comparison, are not derived from the article of Rumpel just named, nor from other treatises on this sub- ject, but are the result of an independent investigation of the older comic poet's 8,835 spoken trimeters. I am greatly indebted to Dr. Arthur A. Bryant for friendly and efficient aid in collecting and tabulating the facts in both Aristophanes and Menander. [Classical Philology IV, April, 1909] 139 140 John Williams White foot has iambic form except the last. There are five non-iambic trimeters in Menander (1 in 145), 1 187 in Aristophanes (1 in 47). 2 Irrational arses preponderate. In the 728 trimeters of Menan- der now under consideration, 199 trimeters (131 with at least one trisyllabic foot somewhere in the verse and 68 that consist solely of dissyllabic feet) have one long arsis (1 in 3.66); 327 (211 + 116) have two (1 in 2.23); 149 (89 + 60) have three (1 in 4.89). The total number of trimeters in which one or more arses are irrational is 675. 3 The total number of irrational feet in the 728 verses of Menander is 1,300, and these are distributed as follows: i* t iii t V * 358 67 49.2 9.2 400 64 54.9 8.8 376 35 51.6+ 4.8-- 'Dactylic' feet Total irrational feet 425 58.4 464 63.7 411 56.45 *Feet are indicated by the lower case numerals : i, ii, iii, iv, v, vi. In the Acharnians, in 811 non- lyrical trimeters, 212 (147 + 65) have one long arsis (1 in 3.83); 325 (202 + 123) have two (1 in 2.5); 220 (131 + 89) have three (1 in 3.69). The total number of trimeters in this play with one or more irrational arses is 757. The total number of irrational feet in the 811 trimeters in the Acharnians is 1,523, distributed as follows: i $ iii i V i 461 31 56.9 3.8 477 65 58.8 8 478 11 58.9+ 1 35 492 60.7 542 66.8 489 60.3 UI. 80, 241, 298; III. 106; IV. 170. 2 Such trimeters do not occur in Aeschylus or Sophocles and only rarely in the later plays of Euripides (Rumpel, op. cit., p. 602). 3 The irrational arsis is allowed, of course, only in the first half of each meter. The spondee appears, however, through haste or oversight, in the second half of the first two meters in several attempted restorations of the broken lines of Menander. The Iambic Trimeter ik Menander 141 Resolved feet (tribrach and 'dactyl') occur on the average in every other trimeter, but the resolutions in Menander (1 in 1.88 trimeters) exceed those in Aristophanes (1 in 2.14). 1 The distribution of the various forms of the foot that are found in the trimeters of Menander is as follows: i ii iii iv V vi 175 35 358 67 93 586 77 '65 216 28 400 64 20 606 71 'hi 279 11 376 35 27 728 Dactyl Anapaest 728 728 728 728 728 728 By resolution of the theses of iambs in the trimeter, the tri- brach may occur in any of the first five feet. The following table exhibits the facts in Menander and Aristophanes: Trim. w^w i ii iii iv V 728 222 35= 18+17 77= 27+50 28= 8+20 71 = 12+59 11= 0+11 8,835 2,654 208= 117+91 960= 384+576 308= 32+276 1,107= 304+803 71= 20+51 711 212 13= 8+5 80= 41+39 25= 4+21 90 = 23+67 4= 3+1 Total Men Ar Lys 222= 65+157 2,654= 857+1,797 212= 79+133 There are, namely, 222 tribrachs in the 728 trimeters of Menan- der: 35 in i (18 contained in a single word, 17 in two or more words or parts of words), 77 in ii, 28 in iii, 71 in iv, and 11 in v. The proportion of tribrachs contained in a single word is slightly greater in Aristophanes (32.3 per cent.) than in Menander (29.3 per cent.). The Nubes is the play of Aristophanes in which the total number of tribrachs is smallest relatively to the number of trimeters (207 in 758) ; the Aves contains relatively the most (331 in 925) ; the Lysistrata is sixth in order. 1 In Sophocles 1 in 16 ; in Aeschylus 1 in 13, in Euripides 1 in 4.65 (Rumpel, op. cit., p. 603). In the Aves, the comedy of Aristophanes that contains the most verses of this sort, 1 in 1.84 ; in the Thesmophoriazusae, which has the fewest, 1 in 2.47. 142 John Williams White The tribrachs in Menander (1 in 3.28 trimeters) slightly out- number proportionally those in Aristophanes (1 in 3.33). The difference in distribution is marked in i and iv. In Menander the tribrachs in i outnumber those in Aristophanes (the divisor 1 is 12) in the ratio of 2 to 1, but in iv, the number in Menander is rela- tively smaller (about 7 to 9). The tendency in Menander to increase in the number of trisyllabic feet in i relatively to those in Aristophanes holds for the- dactyl and anapaest as well as for the tribrach. The following tables exhibit the facts for the tri- brach in Menander in detail: TRIBRACHS CONTAINED IN ONE WORD i ii iii iv T ii 3 2 7 14 8 3 9 4 1 3 18 27 8 12 Total Trisyllable Overlaps forward Overlaps back Overlaps both ways . . . Total in one word 16 32 13 4 65 TRIBRACHS CONTAINED IN TWO OR MORE WORDS OR PARTS OF WORDS i ii iii iv V Total 12 5 4 43 3 5 15 10 49 2 9 21 128 8 17 50 20 59 11 157 In order to facilitate comparison, I add the corresponding tables of the tribrach in Aristophanes: TRIBRACHS CONTAINED IN ONE WORD Trisyllable Overlaps forward. . . Overlaps back Overlaps both ways . Total in one word. 50 67 117 69 190 104 21 384 19 3 2 32 97 79 110 18 304 5 5 10 20 Total 229 360 227 41 857 1 Determined by the ratio of tribrachs in Aristophanes to those in Menander. The Iambic Trimeter in Menander 143 TRIBRACHS CONTAINED IN TWO OR MORE WORDS OR PARTS OP WORDS i ii iii iv V Total In three words 15 56 20 81 462 33 55 219 2 98 676 29 4 46 1 253 In two, divided ^ [ ~w 1459 In two, divided wjv 85 Total divided tribrachs 91 576 276 803 51 1797 Menander excludes the trisyllabic tribrach (|w~~|) from iii and v and the instances of its occurrence in these feet are very rare in Aristophanes. In IV. 308* the Cairo Manuscript of Menander reads: ov8e[v. t( ov]i> evye<> ovtco?, aftekTepe; The verse is unmetrical. It has been emended by transposition of feet in the second meter (ovtg>9 eifio<; oS' 6'itco? kcrri ? e%et?" III. 149*; " aapK 1 , aXX 1 wira'f e? /co/3a/ca?,07rft)?eto-«/a' 67a)" IV. 477*. Menander uses the trisyllabic tribrach twice in iv (IV. 75, 435). 3 Some scholars apparently find warrant in this and in its free use by Aristophanes for employing it in the emendation of Menander, but it certainly should not be allowed to supplant an unobjection- able reading established by the MS. Cf. "It' ova\ a tot : eKeivy 1 See A. Korte in Berichte d. Kon. Sack. Gesell. d. Wiss. LX (1908), III, p. 132. 2 The trimeter has been thus restored by Robert and Croiset. Von Arnim converts the tribrach into an anapaest (irtpvinv vapTjuBa ^uerd xipijs), but there are only three trimeters in all Greek comedy (some 16,000 verses) that begin with three anapaests: Aristoph. Ach. 536; Antiph. 20. 5; Sosipat. 1. 33.— For another doubtful example of the trisyllabic tribrach in iii in Menander, see frag. 481, 5. "Of. also frag. 393, 4; 482, 7; 663, 2. 144 John Williams White yeyovev, airavr'' e/ta" II. 301*. Here the MS reads: err' oixra, to, t' itceiW) 1 yevofteva train 1 e/ta. Menander uses the trisyllabic tribrach four times in ii (II. 425* ; IV. 76, 79, 150). To these examples, which are supported by the practice of Aristophanes, IV. 118* should probably be added: e/toi r e7reo€T av acr/tefo?; wvi, be /tot. Forty-nine of the 222 tribrachs found in unbroken trimeters in Menander are contained in words of four or more syllables, which therefore overlap the neighboring feet. Thirty-two begin with the word, which therefore overlaps the following foot (~~~+); 13 end with the word, which therefore begins in the preceding foot ( + w~w ) ; 4 are contained in words which begin in the preced- ing and end in the following foot (+~~~ + ). No 'overlapping' tribrach is found in v, 3 and its use in an emended line is dubious. Cf. ei fiev jap rj /3oi/Xo'/ieM>? $ KaTe^o'/tevo? " IV. 115; " KcucoSal/wt? ovtcd- Sie^vy', oiiS' evuirviov J7P" IV. 440*. Menander uses the tribrach contained in a word that overlaps forward ( — + ) freely in iv. Cf. II. 147, 198, 501; II. 301*, 436* ; III. 3*, 162* ; IV. 169*. He allows it freely also in iii and is here in striking divergence from Aristophanes, who preserves the penthemimeral caesura. Cf. II. 64, 80, 194; III. 25, 48, 80; IV. 272, 277. In all these cases in iii the word with which the tribrach begins is a quadrisyllable, so that there is division after the arsis of the fourth foot. Tribrachs contained in a word that overlaps both ways ( + — + ) are found in ii in II. 37, 368*, 518 + frag. T„ 1*, and in iv in II. 8, 288, 300. All these verses have penthemimeral caesura. This tribrach is not found in iii, 4 and the following emendation therefore is doubtful: "edvov eVXeXv/ieVo?- aWa firjv 7roeti/" IV. 529 + 519*. The alternative proposal, e/eXeX^oytew)?, is equally objectionable on another account (see p. 154), and the true reading probably remains to be found. The MS according to KOrte 5 has tov vbv e'« . . . . fievos. J K6rte, op. cit, p. 128. 2 K6rte, op. cit., p. 117. This tribrach is not found in the fragments of Menander. 3 It occurs only twice in the fragments : 123, 2 ; 325, 15. 4 It occurs in no fragment. 6 KOrte, op. cit., p. 140. The Iambic Teimbtee in Menandee 145 Finally, to regard the facts in sum (see the tables), Menander in i uses relatively twice the number of tribrachs contained in one word that are found in Aristophanes ; in iv this relation is reversed. Both poets show predilection for tribrachs contained in two or more words or parts of words. The parts of a divided tribrach may be connected with one another within the tribrach by elision, as tV cnroBw II. 148 ; Xey' o Xeyeis II. 298 ; evSov tV avair-qhriaas IV. 438, and either the first or last part may be an enclitic, as el p,r\ ti KaratreircoKe II. 151; avdj3a" IV. 464*. In v: "ei^citf' oV efyalveff rjviica TXWpa" III. 110. Menander uses this tribrach 10 times (twice in broken lines) in the first meter. Its first two syllables consist of an unelided dissyllabic word in common use. Cf. II. 8, 305, 341, and 523*; IV. 316, 321 (the prepositions wept, hri, Kara, fiexpi, irapd) ; II. 330 (tivo? eveicev), 369* (irJdep e X «?) ; IV. 271 (roVe fiev), 313 (a' 6 n fia^a(pa<: irepi $e/3et? • licavbs yap el IV. 69*.' He has it in v in II. 46, 396; III. I*. 2 Aristophanes evidently regarded this tribrach with disfavor in v, since he has it there only four times. It has been used in v in restoring the following lines: "ti yap Tt? av vo/itcreiev aXKo to yeyopl;vfifiepo<; apa" IV. 276*. The single example of the dactyl overlapping both the adjacent feet is found in iii (I. 55), with strong hephthemimeral caesura. 3 Both poets employ the dactyl composed of two or more words or parts of words much oftener than that contained in a single word. Aristophanes especially inclines to its use. The parts of the dactyl may be connected with one another within the dactyl by elision, as ravr' eTTCTidrjvai II. 88; olB' on II. 503; ov% iV aSi- KTjaca II. 373; /care/SaM/' a' wrepwov IV. 17; and either the first or last part may be an enclitic, as -rrptirepSv fioi fiera/ieXei II. 205; el TTore III. 18 ; el 8d ye II. 117. 1 This dactyl is not found in v in the fragments. 2 It is not found in the fragments of Menander. 3 Also in the fragments in iii. Of. 540, 3 ; 1089, 1 ; 1102, 1. The Iambic Tbimetee in Menandeb 149 The prevailing form is the dactyl composed of two words with the division between arsis and thesis (-|--). The order, determined by frequency of use, is here practically the same in both poets, iii, i, v, but Aristophanes prefers this dactyl in iii, thereby again favoring penthemimeral caesura. Except in i, the arsis of the dactyl is generally the final syllable of a word of two or more syllables (cf. the tribrach, p. 145), and arsis and thesis are often separated by strong punctuation. In v this dactyl (-|~w) is frequently so composed that the trimeter ends in a tetrasyllable word (-ww«). Ci. I. 54; II. 42, 56, 88, 135, 165, 205, 241, 286, 318, 465, 488; III. 10, 52, 100; IV. 9, 60, 437 (18 instances in 28 ; in Aristophanes, 93 in 134) . Menander uses the dactyl composed of 3 words freely, even in i. 1 This is at variance with his use of the corresponding forms of the tribrach (p. 142) and anapaest (p. 151) in the same place. In 15 of the 20 instances (excluding those in i) of Menander's use of this dactyl (2 in broken lines), the arsis is the final syllable of a word of two or more syllables. Cf . the corresponding use of the tribrach (p. 145). The dactyl consisting of three monosyl- lables is rare and is found only in II. 92*, 359, 503, all in iii. 2 Both poets deviate from tragic usage in allowing a division of the dactyl composed of two words after the first short syllable of the thesis (-- 1 w), and Menander uses this dactyl in i with great freedom. Both poets exclude it from v, 3 and it should not there be employed in restoring broken lines. Cf. " appev reKovaa irdp- devos 6rj\v ti 0' a/tta" I. 2* ; " erepov ti irpo<; tovtois kvk&v ov/cer' av e^ot" II. 211*; " ToiracrnKov to yvvaiop, a>? ip7re0'; on ycip" II. 340*; " . . . . ak\o yeyovev" II. 387*. Menander has it once in iii, 11 times in i. This dactyl (-~ | w), in both Menander and Aristophanes (53 occurrences), is always followed by an iamb, never by a trisyllabic foot (tribrach or anapaest). The anapaest, although it is an irregular form, is freely used in comedy in the first five places of the trimeter. 4 It is the char- 1 It occurs in i twenty-five times in the fragments. 2 It occurs twice in the fragments : 257, 3 (i), 740 (ii). 3 It is not found in v in the fragments of Menander. 4 Neither the anapaest nor the tribrach ever occurs in vi, of which the dissyllabic form is absolutely fixed. It is a grave mistake thus to restore any verse in Menander, 150 John Williams White acteristic feature of the comic non-melic trimeter, and both in Aristophanes and in Menander anapaests outnumber both tribrachs and dactyls. The use of this form of foot is so natural that in 353 instances in Aristophanes and 32 in Menander anapaestic scansion is secured by position, the original form being a tribrach. Cf. I. 38 (aica/cov), 44 (fteipdiuov) , 59 (e\eu#e/3io?) ; II. 4 (utapcfc), 42 {irpoTepov) , etc. The following table is a summary statement of the use of the anapaest in Menander and Aristophanes: Trim. WW- i ii iii iv V Total Men. .. 728 256 93= 65+28 65= 47+18 20= 13+7 51= 40+11 27= 24+3 256= 189+67 Ar 8,835 3,840 1,158= 495+663 1,209= 928+281 261= 180+81 865= 699+166 348= 260+88 3840= 2,562+1,278 Vesp. . . 752 324 108= 44+64 107= 86+21 21= 18+3 68= 59+9 20= 17+3 324= 224+100 The anapaests in Menander (1 in 2.84 trimeters) are not quite so numerous proportionally as those in Aristophanes (1 in 2.30), but in i they outnumber those in Aristophanes (the divisor is 15) in the ratio of 93 to 77; in ii this relation is reversed (65 to 81) ; in the remaining feet the order is the same in both poets (iv, v, iii). The following tables exhibit the facts for the anapaest in Menander in detail: ANAPAESTS CONTAINED IN ONE WORD i ii iii iv V 36 14 5 18 9 29 23 3 12 14 9 2 7 1 1 3 3 65 47 13 40 24 Total Trisyllable Overlaps forward Overlaps back Overlaps both ways. . Total in one word 82 81 19 7 189 as " airis iripeaTiv oirroal- tJ iraiS&ptov" II. 85* ; " viou 84 • i] Si ko.1 t6 iraiSdpiov ?\a/3ev" II. 433*. Aristoph. Ran. 1203, an intentional vagary, sounds the note of the impending metrical jest. The Iambic Trimeter in Menander 151 ANAPAESTS CONTAINED IN TWO OR MORE WORDS OR PARTS OP WORDS i ji iii iv V Total In three words 1 17 10 7 11 2 5 2 2 7 1 2 4 30 33 28 18 7 11 3 67 The corresponding table in Aristophanes is as follows: ANAPAESTS CONTAINED IN ONE WORD i ii iii iv V Total Trisyllable 279 216 276 419 164 69 46 91 27 16 270 156 230 42 48 145 44 24 919 1,027 465 151 495 928 180 698 261 2,562 ANAPAESTS CONTAINED IN TWO OR MORE WORDS OR PARTS OF WORDS i ii iii iv V Total 114 447 102 31 162 87 7 50 24 23 84 59 7 56 25 182 799 297 663 280 81 166 88 1,278 The three examples in Menander of an anapaest contained in one word that begins with iii and overlaps iv (II. 163, 308; IV. 284) are supported by two in broken lines (III. 34*, 74*). 1 Aristophanes has this anapaest 91 times. The two examples in iii of the anapaest contained in one word that begins in ii (II. 513; IV. 140) are not supported by other examples in broken lines, 2 and 1 It occurs also in frag. 8, 2 ; 74, 7 ; 402, 15 ; 501, 2 ; 542, 4 ; 585, 1. 2 It has the support of four verses in frag. 272, 1 ; 360, 1 ; 393, 2 ; 403, 5. 152 John Williams White it is doubtful whether this form should be introduced, by substi- tution, into a line that is in itself metrically unobjectionable: "/SXei/ra? iXev0ept.6v ti To\fitfe\ev" II. 441*; "Orjiprjfievov eariv; — teal crbv ofioUw. — Tla|Modvii;a in Eq. 1,225. 5 Menander's sparing use of ~ | w | _ has been remarked. Yet editors have used the anapaest contained in 3 words in restoring his text, even in ii and v, where it has but meagre support in the 1 If the enclitic is regarded as inherent part of a combination, we have an instance of an anapaest in ii overlapping both ways. 2 See p. 144. 8 It occurs in v in frag. 540, 4. *It occurs 10 times in the fragments, in 1,470 trimeters. Five of these cases are in i, one in ii, three in iii, and one in v. * An enclitic is found as second word in the fragments of Menander in 223, 3, 6 ti h.v; in 355, 2, tI tot' iarlir, and in 587, 2, 6 re itXoCtoi. The Iambic Tbimeteb in Menandee 155 poet's own practice. Cf. ^irai^eiv, rt? 6 \?}po?; Kara Kpdro<; to Su<7Tv^e5" IV. 469*; "7rat&>?, rl tovtov Bel \a0eiv to ye abv /4>o?" II. 279*; "a^itar' egco tj?e5 dv ti SeiKvvoi rmfir)piov. Here the MS has tovtov aa(f>e<; av AG IK NY . . ANTITGKMHPION. This is unmetrical, but it apparently fur- nishes an instance of the occurrence of this anapaest in v. 1 In III. Ill * the MS has: Xa/3?7 ti tovtcov • ov yap empaicev . . . . , where doubtless, quite apart from considerations of meter, we should read iopaicev, a reading proposed independently by different scholars. In the other three cases correction is not so obviously needed. In the first the anapaest occurs in iv: aXX' awro'Sos el firj — (f)T)o-tv — dpecrKei- tovto yap (II. 129*), in which the reading cf>r)a-' has been proposed (Leo). But (pqo-iv is parenthetical and is intentionally misplaced, in imitation of the disjointed language of excited persons in the street. Khythm cannot be predicated of such a verse. There are two occurrences of the anapaest in ii. Both the verses in which it is found have penthemimeral caesura and the words constituting the anapaest are closely connected. In II. 131 the MS has owe ean Sticaiov, ei ti t&v tovtov ae Set. Here Leo proposes e? to Siicaiov. In III. 77 we read: ovSev ti toiovt' rjv, & IloXefiav, olov cf>aTe. Here it would be possible to delete 3>. In Menander toiovtos is either an amphibrach or an antibacchius (for the latter cf. II. 153*; IV. 160*, 282) and, as is well known, Menander inclines to diaeresis of the trimeter (cf. 1. 24, 38, 53, 60; II. 9, 21, 26, 43, 50, 55, 62, etc.). But the change does not commend itself as necessary. Some editors have employed the anapaest now under considera- tion in emending and completing the trimeters of Menander, but this is not justified in v and iii, by either his own practice or that of Aristophanes. Cf. in v: "fjv koivo<; 'E/^m}?, to fiev av ovroi i\dfi/3avev" II. 67; "owe rfkOov aXXta?. irpoo-fieveiv ti So/cels av /not" II. 364*; "i8ta>T\ — iydr, — So/eet? 76- fiaTrjv Se Xeya) Xofyov 1 ?" (iv, v) IV. 71*. In iii: "7ra? dv ti? erepa crvvfiKe. — TrdvSeivov Xe'yei?" II. 515*; "e7r' dpunepd Towra crvvfiKe. irdvheivov Xeyet?" (+^w|_ also in ii) II. 515*; "dpiaTOV dpicrTOv. — del /j,eva> Tpio-- a0Xto? (ii, iii) T 2 , 2 + 11. 527*; "ai'Ke8' o5to ? - iya> ydp foov ov v\d^ei<; avrb<; direp crvXev; 7' ifi^" II. 157*; "vfilv erepov (lev, /irj /ie KaXelr' • ei? fia/capias" II. 531*; "rovfiov •jrarpb'; koX /iTjTpd'i, e/3ovX6/ivv e%eiv" III. 143*. In ii: " ffkeifrov S^,irdrep, «d«etcr'* IcroK ecrr' ouTOtrt" II. 103; "rfc, ay\ dv n Por)0rjo-eiev, dW $ yeyov' ep5>" II. 387*; "ex' ovcra Tidrjvrjcreo . . . . " IV. 34*; "ti? §' iariv; 1 — er)v Be iravra raXka \av0dveiv" IV. 105*. This anapaest when composed of three words ( + ~ | ~|i - ) is practically barred. It will be remembered that Menander is in general disinclined to use the anapaest consisting of three words (p. 154). The following proposed restorations have little sup- port in Menander's own practice: " iroip,rjv Be Tt? e^rjyyeiXe fioi, 7T/30? 8i> ovrocrV II. 82; "«? icrTi aov, ovk dWoTpiov. ei, yap eoe- Xei»" II. 441*; "eBodv Be t60' ovtco? • yevofie'vcov 8' ir&v nvcov" III. 4*; "tovtwv ere /lev ovBev, co? eyp/Aai, Xavddvei" IV. 83*; "ovk ovra t' iv avrov- iroWd Be o-vvepyd^eTai" IV. 125*; " Kavravda crv fir) pevcov to yeyovos Kpvfyd' 1 otroi'" IV. 136*; "7rovrjpb<; 'qcrOa — ere 7' & irovrip' • 07T(»9 crv vvv" II. 156*; "eyeb Kara iroWo 7c vvv (lev ovv, ovk oZS' 07r«9" T 2 , 3 + II. 528*; "^rjTOVfiev Xerov ovr\ el Be ere fivSev «co\v«" II. 10*; "ra %pverp rj ravff'' arret, 7tot' ecrri, irorepa Bel" II. 92*; "aim', oil yap av iJKOifi' av0K — dWd ri y" IV. 464*; " vofiio-aa 1 ave/cpaye ical Oepairaivihicp tivV IV. 36*. Only the first of the particular combinations used in these three restored trimeters is found anywhere in Greek comedy, but in this case (III. 31*) the correction of vttotov, the reading of the MS, is obviously not virb tovtov, but, as Van Leeuwen has seen, inr' avrov. The famous forbidden combinations, -~~~~- and ~~~~w-, which have been the subject of long-continued discussion, 2 are found in the Cairo MS of Menander, but in trimeters that arouse suspicion because they are objectionable on other accounts. The first combination thus occurs three times. In II. 451* the MS has a\\a TT6PIM6NQ KaTaayelv ttjv -lrpolica /iov. The verse not only begins with a meter of the tabooed form, but is also unmet- rical in the last two meters. Here irept,p,eivm (Cronert, Headlam, 1 Three combinations that do not occur in the table are found in broken lines : ~ — ~~_ iniii, v (III. 34*); -~~ -„~ in iii, v (IV. 130*) ; ~~- w~^ -wim, i, iii (IV. 69*). These are all found in Aristophanes also. 2 See Dawes Miscellanea Grit., p. 253: "I. Dactylo anapaestus subjici nequit. II. Neque tribrachyn sequi potest anapaestus." Hermann Elem. Doctrinae Met. 126-39. Reisig Coni. 11-86; Dobree Addenda to Porson's Aristophanica, pp. 111-15 (ed. Wagner, Observationes, pp. 22-27); Rossbach and Westphal MetrOc* 227-30; Christ MetHk" 328. 160 John Williams White Van Herwerden) restores a perfect trimeter. In II. 508* the MS has ATTANTATArAOA — tI Xva irapi-^m for which there is no reason- able explanation, and either hiatus and the forbidden combination or an intolerable synizesis. Lefebvre had previously proposed ij /ioi So? avrbv Xva trapiyja aS>v — fiovko/xai, which gives the second objectionable combination. Various restorations have been pro- posed. Granted that the order of the words has been disturbed, as the position of avra> would indicate, it is possible that Menander wrote v\ fioi So? Xva cr&v avrbv Trape^ca. — ftovXofiai. Some trimeters have been restored with the objectionable combination: "%p6vov, SiaKwrrtov eviore, iravra^ol 07CO7r" II. 389*; "^/seoTce? axny TcL^a •nysoVepo? vxiv S' ov/cert" III. 82*; iraiBiov iiceivov yeyove, rk aXkrj ical r48e" IV. 33*. The other combination (~ ~ ~ ~ ■ — ) is found in the second meter of IV. 192*: dWrj- Tt 7tot' io~Ti to yeyovos; — e/c/8e/3Xj;«e p,e, with an anapaest of doubtful form in iv (p. 153). By omission of eo-rt (Leo) the trimeter is restored to proper form, dWrj. Tiirore to yeyovosf This combination is found also in some restored trimeters: "to ye'vos inrep r)/j,a<}, ical Tpacfrels iv e/J7aTat?" II. 104*; "7rd? dv rt? erepa crvvfjice. — irdvBeivov \eyet?" II. 515*; "on OvyaTpiSoik o-oi yeyove; 7ra)?; vt) tov? #eov?" IV. 528*. My purpose in publishing the preceding statement of facts is entirely practical. A happy fortune has recently given us con- siderable parts of four comedies of Menander. These plays are simple and charming and are likely to be widely read. It is certain that many editions of them will appear. These comedies should go hand in hand with those of Terence and should serve as an introduction to the plays of Aristophanes. 1 Korte, op. cit., p. 127. 2 Cf. frag. 248, 5, ot^/Soi/Xos- 6 Suvptpav Xoyiff/up wiyr' «x«. where Porson proposed 6 \oyuriMp 5iaip*j>v; 384, paicdpios 4neivos, where Kock proposes ndicap y' (kcivos; and 663. 3, t& aKdifip.ad'' ola ret ao