^{page 152 note 1} Accentua Symmetry in Vergil (pp. 10–13).

^{page 152 note 2} Cf. De Rerum Natura 1. 6 and 1. 66 for specia emphasis obtained by homodyne in each of the first two feet (this is not usua). References throughout are to Munro's edition of 1886; when this work was begun, Bailey's edition was not yet availalbe in Australia.

^{page 152 note 3} I have assumed that a monosyllable can carry a stress accent. I consider that elision does not cause any change in accent; also that enclitics (-que and -ve) so coalesce as to shift the accent. For examp e, in 4. 740, the elision of the a of natura does not change the accent; I consider it justified to assume that the fifth foot is homodyned; hence the line ends natúr (a) animántis. Similarly in, say, 4. 556, on the justifiab e assumption that in these frequent cases the fifth foot is homodyned as usual, the line ends servátque figúram.

^{page 154 note 1} As a general rule; see Bailey, p.116.

^{page 153 note 2} Op. cit., pp. 40, 41.

^{page 154 note 3} Op. cit., p. 95.

^{page 154 note 4} The number of divided ines, however, is perhaps too small to warrant aconc usion.

^{page 154 note 5} Books 3 and 6 appear exceptiona. In Book 3, 8 out of the 16 divided ines are heterodyned. But in 5 of these lines Lucretius is satisfied with a prosaic line, using the word enim to connect the two sentences; e.g. lines 365, 731, 881, … neque enim; line 162, … ubi enim; line 6,… quid enim. Similarly in Book 6, 2 of the 6 examples of heterodyned divided lines use enim after the caesura; line 133, … etenim; line 263, …neque enim. This use of enim automatically makes the fourth foot heterodyned. In such ines it appears that ucretius was satisfied to introduce the new sentence by the weak and prosaic enim rather than to use the effective device which he preferred in most divided lines.

It is interesting to note that the order of the books in descending order of percentage of homodyned ines, from the above table, is 1, 2, 5, 4, 6, 3. Bailey gives the order of composition as 1, 2, 5, 4, 3, 6. The c ose similarity suggests that Lucretius either became progressive y less careful in using homodyne for this purpose, or, in the later books, was more conscious of the necessity of avoiding the monotonous sing-song of homodyne rhythm.

^{page 154 note 6} I have found alternations within the following passages: Book 1: 1–7, 288–94, 298–303. 329–35. 398–403, 440–8, 464–8, 551–5, 561–9, 612–22, 683–9, 851–6, 861–6, 861–6, 897–903, 936–42, 977–83, 988–93, 994–8, 1008–13, 1061–7.

Book 2: 14–19, 66–71, 91–99, 106–11, 188–93, 256–60, 314–20, 333–7, 360–4, 384– 9, 391–7, 414–18, 422–7, 464–8, 517–21, 592–7, 661–6, 669–76, 707–11, 741–5, 768–77, 851–5, 868–72, 966–70, 973–7, 1004–8, 1077–83, 1119–23.

Book 3: 23–27, 69–75, 84–88, 121–7, 231–40, 252–6, 337–43, 437–43, 462–6, 481–94, 502–8, 598–602, 702–7, 717–21, 787–93, 847–51, 1003–11, 1014–22, 1024–8, 1063–9.

Book4: 11–17, 42–47, 81–86, 193–8, 203–8; 221–6, 244–9, 420–5, 438–42, 580–4, 598–602, 655–60, 845–9, 907–13, 929–36, 991–7, 1094–1100, 1112–16 1129–36, 1155–9, 1173–9, 1218–24.

Book 5: 16–21, 41–46, 59–64, 65–70, 131–7, 308–13, 345–50, 402–10, 489–94, 568–74, 579–84, 602–7, 643–9, 776–80, 788–96, 824–9, 849–56, 864–9, 878–84, 988–93, 1034–40, 1043–7, 1057–63, 1113–17, 1119–23, 1281–6, 1379–83, 1431–5.

Book 6: 1–12, 43–47, 50–57, 125–9, 145–53, 206–10, 300–7, 348–54, 368–74, 379–83, 417–22, 474–82, 489–94, 515–21, 540–9, 633–7, 641–6, 658–64, 680–4; 743–7, 748–52, 830–5, 859–68, 927–32, 1053–7, 1058–62, 1098–1102, 1168–75, 1190–6, 1213–18.

^{page 155 note 1} 1–7; 2. 14–19; 3. 23–27, 69–75, 84–88; 4. 11–17; 5. 16–21, 41–46; 6. 1–12, 50–57. This seems to indicate that Lucretius favoured alternation in careful composition.

^{page 155 note 2} I use ‘a’ to denote a line heterodyned in fourth foot, ‘b’ to denote one homodyned he fourth foot.

^{page 155 note 3} For released movements see p. 158.

^{page 156 note 1} Just as Lucretius used alternation in each proemium so also he uses it in the opening lines of important sections of the poem; e.g. 5. 55–7; 1. 329–35, 1008–13; 2.333–7; 4. 907–15. 929–36; 5. 1281–6; 6. 680–4.

^{page 156 note 2} Another passage in alternation in this important section is 3. 1063–70. Again, 1. 93–100 is a passage of ‘obviously intentional solemnity’ (Bailey, p. 77) in which alternation occurs, ine passage (describing the sacrifice of Iphigeneia at Aulis) shows careful metrical construction producing a balanced rhythm ba ba bb aa.

^{page 156 note 3} See p. 120.

^{page 156 note 4} Examples are 1. 288–94, 1029–34; 2. 314–20, 592–7. 661–6; 3. 998–1028; 4. 581–4, 438–42, 991–7; 5. 402–6; 6. 145–53. 641–6.

^{page 158 note 1} When these lines are read the alternation seems more obvious to the ear than usual. Since each fourth foot is a spondee the difference in die position of the stress accent is more noticeable.

^{page 158 note 2} Well shown in line 490 desipit, extentat nervos, torquetur, anhelat.

^{page 158 note 3} Other examples are 2. 741–5; 4. 1218–24; 3. 121–7; 5. 788–96; a so 1. 398–403, 464–8, 551–5, 683–9, 851–6, 891–6, 988–93, 1061–7; 2. 517–21; 4. 843–50; 5. 1113–26, 490–4. Rhetorica questions or supposed objections are emphasized in mis way in 1. 897–903; 3. 121–7, 437–43, 717–21; 5. 308–13; 6. 417–22.

^{page 158 note 4} 3. 231–40, especially lines 231, 234, 235; 3. 337–43. especially 337–8; 3. 787–93; 5. 1034–40, 878–84

^{page 159 note 1} Examp es of re eased movements occur in:

Book 1: 17–20, 151–4, 161–4, 238–40, 252–8, 259–64, 302–4, 325–7, 370–4, 375–80, 426–9, 528–31, 722–5, 742–5, 794–7, 851–3, 927–30, 943–6, 947–50.

Book 2: 184–7, 230–4, 325–30, 398–401, 426–30, 515–18, 600–3, 640–3, 980–4, 1019–22, 1090–2, 1109–11, 1150–2.

Book 3: 94–97, 130–4, 203–5, 288–93, 391–5, 403–5, 421–4, 470–3, 495–8, 591–4, 664–6, 730–4, 888–93, 935–9, 1036–41.

Book 4: 1–5 (= 1. 926–30), 18–21, 22–25 (= 1. 947–50), 95–97, 129–35, 161–5, 256–8, 275–8, 304–7, 318–23, 325–8, 380–5, 386–8, 389–94, 395–9, 482–4, 494–8, 584–9, 636–9, 642–4, 677–83, 684–6, 739–43, 916–19, 929–31, 962–5, 1091–6, 1117–20, 1192–4, 1197–200.

Book 5: 1–5, 39–42, 148–51, 200–3, 257–60, 318–21, 539–44, 561–3, 566–9, 575–7, 592–5, 705–12, 727–30, 828–31, 933–6, 943–7, 948–52, 958–61, 1091–3, 1169–71, 1194–7, 1341–6, 1357–60, 1361–4, 1384–7.

Book 6: 13–19, 108–12, 156–9, 167–72, 209–16, 236–8, 455–8, 465–7, 513–16, 608–10, 631–4, 998–1001, 1005–8, 1062–4, 1103–5, 1158–62, 1178–81, 1230–4, 1252–5, 1276–81, 1282–6.

^{page 159 note 2} Book 1: 120–6, 471–7, 478–82, 628–34.

Book 2: 297–302, 859–64.

Book 3: 18–22, 221–7, 396–401, 634–9, 674–8, 679–84, 748–55, 794–9, 838–42, 870–5.

Book 4: 6–9, 111–15, 400–3, 453–61, 500–6, 544–8, 752–6, 1263–7.

Book 5: 632–6, 751–7, 925–30, 953–7.

Book 6: 189–93, 703–7, 762–6, 1247–51

3 Others can be found above, n. 1.

4 Other examples are 1. 20, 3. 405, 4. 9.

^{page 160 note 1} These lines describe Aetna. The pent-up power of the volcano is imitated in the hetero-dyned verses, and the unhindered bursting of the flames is imitated in the final homodyned line.

^{page 160 note 2} This is a description of legions charging on the plains. The last six lines form a released movement. The effect of the series of heterodyned lines is to suggest the confusion and conflict and powerful movement of the scene as the legions charge; the long descripforth tion is brought to a satisfactory close by an effective homodyned line: tramittunt valido quatientes impete catnpos.

^{page 160 note 3} Op. cit., p. 120.

^{page 161 note 1} Book 1: 80–86, 556–64, 670–9 (10 homodyned verses), 909–14.

Book 2: 142–9, 985–90, 1131–45.

Book 3: 189–95, 447–54, 474–81. 510–18, 560–5, 722–9, 854–69, 904–11, 1080–94.

Book 4: 513–21, 615–24, 1105–14, 1209–17

Book 5: 97–103, 449–57, 680–8, 962–72, 1067–77, 1143–51, 1262–72, 1287–92, 1426–33, 1448–57.

Book 6: 80–91, 180–8, 219–27, 1065–71.

^{page 161 note 2} Exceptional uses of them for this purpose may be found in 1. 80–86; 3. 904–11; 3. 1080–94. These three passages form such a small proportion of all the sequences of homodyned verses that it cannot be claimed that Lucretius consciously used this metrical device for the purpose suggested.

^{page 161 note 3} I have found only one simile so distinguished, and although it is a good example of it, it seems far more likely to be a chance occurrence–4. 513–21.

^{page 161 note 4} Some passages which may be so classified are 5. 962–72, 1143–51; 2. 985–90; 3. 510–18; 5.1426–33, 1448–57.

^{page 161 note 5} Since elsewhere he has spoken of heterodyne expressing reluctance and conflict, one would expect a sequence of heterodyned verses to be appropriate for a conflict or vehement struggle between two forces.

^{page 161 note 6} Book 1: 102–9, 127–35. 149–58, 340–50, 511–19, 635–40, 705–15, 722–30, 734–41, 770–7, 845–52, 823–9, 1074–82.

Book 2: 106–13, 203–9, 261–8, 281–7, 374–80, 547–59, 640–54, 931–6.

Book 3: 307–22 (16 heterodyned verses), 406–16, 526–32, 659–66, 730–40, 978–86.

Book 4: 54–64, 152–65, 279–88, 292–301, 364–74, 557–67, 762–70, 971–7, 1008–96, 1331–40.

Book 5: 26–36, 71–81, 273–80, 999–1008, 1105–12, 1211–17, 1233–40, 1293– 9, 1330–40.

Book 6: 108–15, 269–78, 396–405, 465–72, 577–84, 596–607, 724–31, 818–25, 866–72, 979–94, 1034–41, 1090–7, 1125–32.

^{page 162 note 1} Cf. 3. 978–86; 5. 26–36.

^{page 162 note 2} Cf. 2. 547–59; 5. 999–1006; 1. 823–9; 4. 292–9.

^{page 162 note 3} Cf. 2. 261–8; 2. 374–80; 3. 659–66; 4. 54–64.

^{page 162 note 4} Cf. 5. 71–81; i. 127–35; 5. 1211–7; 4. 1131–40; 6. 596–607; 1. 722–30, 734–41; 1. 635–40. These passages are all more skilfully written than the theoretical and argumentative parts, and it is clear that Lucretius favoured heterodyned verses for such passages.

^{page 163 note 1} Complex Patterns. In chapter vii Knight discusses complex patterns in Virgil. These patterns are formed by combinations of released movements and alternations. On p. 67 he says: ‘These elaborate structures are what I want in particular to exhibit in diis book.’ I find, however, that this part of his book is least satisfactory and least convincing. In some of the longer passages which he quotes Knight appears to have been too ingenious in detecting patterns. Passages which perhaps show no deliberate scheme of construction have been made to appear as ‘elaborate structures’ through a too great readiness to accept most sequences as ‘patterns’. I see no point in an ‘exact returning symmetry’. Why would the poet wish to construct such an elaborate and hidden symmetry? An examination of the poem reveals little evidence that Lucretius consciously designed complex patterns