^{page 454 note 1} The writer wishes to express his warmest thanks to Professor E. H. Wilkins of the University of Chicago, who very kindly read the manuscript of this article and made a number of valuable suggestions.

^{page 454 note 2} All students of early Italian literature appreciate the great service rendered by Biadene's Indice delle canzoni italiane del secolo xiii, Asolo, 1896, and Gnaccarini's Indice delle antiche rime volgari a stampa … della Biblioteca Carducci, Bologna, 1909. Though the present study is the result of an entirely independent examination of the materials, it has frequently profited by these two works. Biadene's index confines itself to the canzoni and does not mention printed editions or metrical forms. It refers to only five manuscripts, the most important ones, to be sure; furthermore, it does not limit itself to the Sicilian group, but embraces the whole 13th century. The Carducci-Gnaccarini index is a remarkably complete list of first lines and printed editions; but in the citations of manuscripts it is very deficient, owing doubtless to the fact that it is a posthumous work. For example: of the thirty-five sonnets in our Repertory, sixteen lack the ms. indications entirely in the Carducci-Gnaccarini, and a number of others are incomplete. Moreover, the order of the references is quite unsystematic, and no indication is given as to whether the pieces are canzoni, sonnets, or ballate, etc. In our list we endeavor to give all the important ms. references, but only a few of the most important and accessible printed forms.

^{page 456 note 1} In cases where different manuscripts attribute one poem to different poets, we find that the latter belong to the same period. Thus we do not find, for example, a canzone of Giacomo da Latino attributed to Guittone di Arezzo or vice versa. From this Monaci concluded that these divergences are due to the original headings of the poems, which named not only the author, but also the person or persons to whom they were dedicated or sent. The copyists of these pieces recorded sometimes the whole heading, sometimes only the poet who wrote it, and sometimes, by mistake, only the person to whom it was sent. Monaci, Sulle divergenze dei canzonieri, in the Rendiconti della R. Accademia dei Lincei, Sept. 6, 1885. This theory should not be overworked.

^{page 456 note 2} In Codex V (Vaticano 3793), “l'ordine delle Canzoni è per Scuole distribuite secondo le rispettive suddivisioni geografiche. Viene prima la Scuola sicula (presa nel suo più largo senso) in cui si veggono in complesso precedere i poeti dell' Isola, quali Giacomo da Lentino (nn. 1-16), Tommaso di Sasso e Guido delle Colonne (nn. 20-23), e a questi tener dietro prima i poeti delle altre provincie del sud, quali Rinaldo d'Aquino (nn. 27-34), Pier delle Vigne (nn. 37-39), Giacomino Pugliese (nn. 55-62), indi i pochi delle altre regioni italiane. Segue la Scuola bolognese rappresentata dal Guinicelli (nn. 104-106), da Nascimbene (n. 101) e da Tommaso da Faenza (nn. 108 e 109), e per ultima la Scuola toscana, etc.” Caix, Origini della lingua poetica italiana, Firenze, 1880, p. 20.

^{page 457 note 1} Reference works and a few historical facts are given under each name in the alphabetical list of poets below.

^{page 457 note 2} Cod. V, 146. It must have been written before 1265.

^{page 457 note 3} Crestom., 216.

^{page 457 note 4} Observe the association of his name with Rosso, who, according to Torraca's proposed identification, belonged to the earliest generation.

^{page 458 note 1} The form of commiato in his canzone (no. 68), not found in the other poets of the earlier period, suggests a relatively late date and the influence of Guittone as far as this poem is concerned.

^{page 458 note 2} To Bonagiunta da Lucca codex P assigns, probably incorrectly, canz. 53, which in V is under Rugieri d'Amici. As Bonagiunta is well known to have flourished immediately after our period, his name is not included here. His poetry is found in A. Parducci, I rimatori lucchesi del sec. xiii, Bergamo, 1905. See also Bertoni, Il Dvecento, pp. 80, 269.

^{page 459 note 1} Studi, pp. 101, 102.

^{page 459 note 2} Scandone proposes Notaio Arrigo di Lentino as the possible author; Notizie, 269.

^{page 460 note 1} The canzone Oi lassa 'namorata, under Odo delle Colonne in D'Ancona e Comparetti's edition of ms. V, is anonymous in Satta (edition of the Società Filologica Romana); but this is probably an error.

^{page 460 note 2} Torraca does not give the list as the one he would finally accept, as later on he questions Inghilfredi and others.

^{page 460 note 3} Crestom., 204. The form of Commiato used by him is that of the Guittonian period.

^{page 460 note 4} Two of these have been assigned, probably correctly, to the year 1262.

^{page 461 note 1} Barbieri, Origine, pp. 143, 145.

^{page 461 note 2} Bertoni, Il Duecento, p. 75, mentions the first two; Scandone, Notizie biogr., p. 324, mentions Arrigo.

^{page 462 note 1} Monaci, Crestom., 271: “La canzone deve essere stata composta poco tempo dopo la battaglia presso Ponte a Valle (25 giugno 1268), che gonfiò di molto l'animo dei ghibellini e li fece confidare fermamente in un finale buon esito.” The metrical scheme we have adopted for this canzone differs from that of D'Ancona e Comparetti and Monaci.

^{page 462 note 2} For a complete bibliography consult G. Bertoni, Il Duecento, pp. 266-271; the bibliographical notes in Gaspary, Storia della lett. ital. and in the manuals of Torraca and D'Ancona e Bacci. Also Gnaccarini, Indice delle ant. rime volg. vol. i, pp. xvii-lxvi; Monaci, Crestom., 697-701; Kritischer Jahresbericht über die Fortschritte der roman. Philologie, München, 1892-, and the Supplementheft zur Zeitschrift für roman. Philologie, Halle, 1877-.

^{page 464 note 1} See also Caix, pp. 19-24.

^{page 465 note 1} Caix, pp. 6-9.

^{page 465 note 2} Caix, pp. 15-18.

^{page 465 note 3} Caix, pp. 30-32.

^{page 466 note 1} Massèra, Zts. f. rom Phil., xxvi, 19 ff.

^{page 466 note 2} Cf. Massèra, op. cit., p. 8.

^{page 466 note 3} Massèra, op. cit., pp. 10-11; Caix, pp. 11-14.

^{page 467 note 1} Massèra, p. 11; Mazzatinti, Inventario dei edd. ital. delle bibl. di Francia, i, 109; Casini in Gior. Stor., iii, 162 and note.

^{page 467 note 2} Cf. also Barbi, Studi di MSS., etc., Bologna, 1900.

^{page 467 note 3} Also called Codex Pucci.

^{page 467 note 4} Also called ‘Testo di Pier del Nero.‘

^{page 467 note 5} See Barbi, op. cit., p. 6, and Massèra, p. 68, note.

^{page 468 note 1} C. 60, 76, 82 and Frag. 2, 3 are probably by Giacomo da Lentino. See Index of First Lines.

^{page 472 note 1} The Istefano di Messina to whom C. 19 is attributed in LR may be another person, a notary of the port of Messina, according to Scandone, Notizie, 356, 357.

^{page 478 note 1} The name added by a later hand.

^{page 486 note 1} The name added by a later hand.

^{page 486 note 2} Satta (Soc. Fil. Rom.), gives it as anonymous.

^{page 487 note 1} Though the poet's name is erased the copyist had written a name beginning Ser guilg …, which was cancelled and Messer Rinaldo daquino substituted. Then this also was erased.

^{page 488 note 1} The name and the first twelve lines were on the lost sheets.

^{page 491 note 1} Most probably by Arrigo Testa d'Arezzo. Cf. Monaci, Sulle divergenze, p. 661, and Biadene, note in the Indice.

^{page 492 note 1} See Cas. Ann.; Flamini, Studi, 179; Torr. 93-96; Bartholomæis, Rime antiche senesi, p. 31.

^{page 492 note 2} See note to no. 2.

^{page 492 note 3} As the quatrains in almost all the sonnets are abab, we give only the form of the sestet for all except nos. 3 and suppl. 3, 4. All the sonnet lines have eleven syllables.

^{page 501 note 1} Cf. Dante, V. E. ii; the studies of Biadene, Stengel, D'Ovidio, Lisio, etc. Bibliography in Bertoni, p. 271. Our metrical schemes differ in a number of cases from those of Lisio.

^{page 501 note 2} Or 6 lines, if the schemes of Casini and Cesareo are accepted for no. 43. This would modify the rest of the above statement, making one canzone with 6 lines and eight canzoni with 8 lines.

^{page 501 note 3} Cf. Dante, V. E. ii, 5 and 12; Blanc. Gram. d. ital. Spr., p. 744; Lisio, p. 31.

^{page 501 note 4} Cf. V. E. ii, 12, § 5, “dummodo in tragico vincat endecasillabum et principiet.” As will be seen in the list the practice of the Sicilian poets is not at all in accord with this rule of Dante's. Dante gives scant consideration to the popular octosyllabic forms, and this accounts for his putting the quinari in the third place.

^{page 501 note 5} Contrast with Lisio, p. 31: “I tipi con quinari, o soli o intrecciati, sono pochissimi.”

^{page 502 note 1} V. E. ii, 5, § 6: “Neasillabum vero, quia triplicatum trisillabum videbatur, vel nunquam in honore fuit, vel propter fastidium obsoluit.”

^{page 502 note 2} V. E. ii, 12, § 5.

^{page 502 note 3} Thirty-three counting Fragment 1 (Allegru cori).

^{page 502 note 4} Cf. V. E. ii, 5 and 12; Lisio, p. 31.

^{page 502 note 5} Or eight, according to Casini's scheme for no. 43.

^{page 502 note 6} This is quite in contradiction to the general statement of Stengel (Grundriss, ii, 1, p. 85): “der Italiener kennt so gut wie gar keine verschiedenversige Strophen, wiche andere als Endecasillabi und Settenarj mit einander verknüpfen.”

^{page 503 note 1} Cf. Dante, V. E. ii, 10; D'Ovidio, Versificazione, 569 ff.

^{page 503 note 2} The term Volta is also very commonly applied to the whole second division of the stanza; cf. Biadene, Varietà, pp. 73, 74.

^{page 503 note 3} Dante, V. E. ii, 12, § 6.

^{page 503 note 4} Nos. 27, 39, 56.

^{page 504 note 1} V. E. ii, 13, § 6.

^{page 504 note 2} “et diesis esse non potest, secundum quod earn appelamus, nisi reiteratio unius ode fiat, vel ante diesim, vel post, vel undique.” V. E. ii, 10, § 3.

^{page 504 note 3} “Et duos (pedes) habere decet, licet quandoque tres fiant: rarissime tarnen,” Dante, V. E. ii, 10, § 3.

^{page 505 note 1} Contrast with Blanc, Gram. d. ital. Sprache, p. 743: “Selten besteht die erste Hälfte nur aus zwei Gliedern von zwei Versen ab, ab.”

^{page 507 note 1} “In versibus quoque fere semper hac lege perfruimur; et ‘fere’ dicimus, quia propter concatenationem prenotam et combinationem desinentiarum ultimarum, quandoque ordinem iam dictum perverti contingit;” (V. E. ii, 13, § 7).

^{page 508 note 1} The lettering is here changed in order to correspond to the piedi types.

^{page 509 note 1} Nos. 3, 12, 13, 14, 15, 22, 27, 31, 33, 36, 38, 39, 42, (43), 47, 49, 51, 56, 59, 61, 68, 78, 81, 85.

^{page 509 note 2} No. 63.

^{page 509 note 3} Cf. Dante, V. E. ii, 10. The combination of Fronte + Coda was not considered possible by Dante.

^{page 509 note 4} See above under ‘Fronte and piedi.‘

^{page 510 note 1} Cf. Monaci: “La canzone del Re Enzo, (our no. 75), … presenta l'unico esempio a me noto di stanza con due chiavi.” (Rendiconti … dei Lincei, 1885, p. 357). Monaci's use of the word chiavi here seems unusual, as the chiave should rime with the corresponding lines in the other stanzas, as in no. 79.

^{page 510 note 2} In no. 69 the chiave, ending in voi or cui, has 9 sylls, not 7, as in D'Anc. C. Ms. Ch. has it correctly indicated.

^{page 510 note 3} Dante, V. E. ii, 13, § 3; Biadene, Collegamento, 8; Stengel, Gr. ii, 1, 83. Biadene proposes that the metaphor in the word chiave suggests the iron link used in architecture to strengthen walls.

^{page 511 note 1} I. e., before the diesis. Compare on the whole subject, Biadene, Collegamento.

^{page 512 note 1} Bartsch, Die Reimkunst der Troubadours, pp. 182-186, (published in Jahrbuch für rom. u. engl. Lit., vol. i, 171 ff.). Biadene, Collegamento, 5, 6. Compare also Petrarch's canzone, “S'i' dissi mai ch'i venga in odio a quella.” For a discussion of imperfect rimes such as mento: neente, intendo: tanto, stringa: segna, tolto: acorto, etc., see Biadene, La rima nella canzone ital. Many of these imperfections are due to dialectical differences, and to carelessness in copying. Sdrucciole rimes occur in only two canzoni of this period and in all cases are verbs from the Latin third conjugation, intendere: rendere (80) and credere: credere (67). Perhaps they should be syncopated, intendre, etc.

^{page 513 note 1} V. E. ii, 5, § 6.

^{page 513 note 2} “Minime autem trisillabum in tragico videtur esse sumendum per se subsistens, etc.” V. E. ii, 12, § 7.

^{page 513 note 3} This modifies Stengel's statement: “In der ital. Lyrik findet sich die rima al mezzo anfangs nur spärlich.” (Gr. ii, i, p. 69.)

^{page 514 note 1} There are also eight in the sonnets, all in one poem.

^{page 514 note 2} It is difficult in many cases to decide whether a line is e. g., 6.5 or 5.6, 6.5 or 7.4, etc.; hence the above figures are given with reserve.

^{page 515 note 1} There is of course the contrary possibility that canzoni with internal rime originally irregular may have become regularized in the process of repetition and copying.

^{page 515 note 2} A practice very common in Provençal poetry, where, however, there was usually in such cases an even number of stanzas; Bartsch, Die Reimkunst der Troubadours, p. 174; Biadene, Collegamento, pp. 6, 7.

^{page 516 note 1} Cf. Stengel, Gr. ii, i, p. 79; Biadene, Collegamento, p. 13.

^{page 516 note 2} Cf. L. Biadene, La Forma metrica del Commiato, pp. 357 ff.; also Stengel, Gr. II, i, p. 83.

^{page 517 note 1} The latter is unissonans and collegate, the former not.

^{page 518 note 1} Cf. Gaspary, pp. 30-33.

^{page 518 note 2} In the above index they are all punctuated cdc, dcd, keeping the division made in the mss.

^{page 519 note 1} Cf. Biadene, Morfologia del sonetto, p. 19, note.