¹ Soubiran, J., Cicéron: Aratea, fragment poétiques (Budé, Paris, 1972), 106–37. As F. R. D. Goodyear points out in his review, CR 92 (1978), 32, the elimination should have been carried through in the apparatus.Google Scholar

² Bischoff, in Karl der Grosse: Werk and Wirkung (Aachen, 1965), 307;Google ScholarSaxl-Meier, , Catalogue of astrological and mythological illuminated manuscripts of the Latin Middle Ages, III (London, 1953), 1,149–51.Google Scholar

³ For reproductions see Ottley, W. Y., Archaeologia 26 (1836),Google Scholar Plate VII opposite p. 149 and Plates VIII, XII, XIV, XVI–XIX, after p. 150; Buescu, ed. (Paris and Bucharest, 1941), Plate I; Saxl-Meier, op. cit. (n. 2), l.xiii, 2, Plates 57, 60–1, 63–4, 66–7; Grabar, A. and Nordenfalk, C., Early Medieval Painting (1957), p. 91 (in colour).Google Scholar

⁴ Buescu, op. cit. (n. 3), 108–9. For a reproduction see ibid., Plate II.

⁵ A full text of these would have to be pieced together from Breysig, , Germanici Caesaris Aratea cum scholiis (Berlin, 1867), 105–232;Google ScholarManitius, , Rh. Mus. 52 (1897), 305–32;Google Scholar and Maass, E., Commentariorum in Aratum reliquiae (Berlin, 1898), pp. 99–312.Google ScholarMartin, J., Histoire du texte des Phénomènes d' Aratos (Paris, 1956), pp. 42–5, provides the best survey of what it would embrace. See also n. 55.Google Scholar

⁶ Winterfeld, P. von, ‘De Germanici codicibus’, Festschrift Johannes Vahlen (Berlin, 1900), 398–9.Google Scholar The objections of Sabbadini, , Riv. Fil. 39 (1911), 244–5Google Scholar = Storia e critica di testi latini (Padua², 1971), pp. 135–6,Google Scholar repeated by Buescu, op. cit. (n. 3), 92–3, rest on the shaky assumption that the Vercellensis read exactly what appears in Oliverius, H., Commentariorum Cyriaci Anconitani nova fragmenta notis illustrata (Pesaro, 1763), p. 43;Google Scholar cf. Traglia, A., Ciceronis poetics fragmenta II (Rome, 1952), pp. 18–21.Google Scholar

⁷ Saxl-Meier, op. cit. (n. 2), xix, xxi, le me to Tietze. I am much obliged to the Warburg Institute for access to a complete set of photographs.

⁸ The script is of a type very hard to date. Cf. Masai, and Wittek, , Manuscrits datés conservés en Belgique II (Brussels, 1972), Pls. 225–6 (1408);Google ScholarSamaran, and Marichal, , Catalogue des manuscrits en écriture latine portant des indications de date, de lieu ou de copiste VI (Paris, 1968), Pl. 88Google Scholar (Milan, 1428–39), II (1962), Pl. 101 (Milan/Pavia, 1442), I (1959), Pl. 115 (Milan, 1459); Lieftinck, G. I., Manuscrits datés conservés dans les Pays-Bas I (Amsterdam, 1964), Pls. 401Google Scholar and 406 (Piedmont, 1440 and 1493); Watson, A. G., Catalogue of dated and datable manuscript. c.700–1600 in the Department of Manuscripts, the British Library (London, 1979), Pl. 794 (Milan, 1477).Google Scholar

⁹ For descriptions of D see Manitius, op. cit. (n. 5), pp. 306–7, and Maass, op. cit. (n. 5), pp. xxi-xxii; Hyginus intervenes between the scholia Sangermanensia and Cicero. G has lost leaves after ff. 3, 9, 14, 16, but as I cannot work out the quiring from the photographs, I do not know whether Hyginus is likely to be missing after f. 16; on other grounds probably not (cf. n. 55).

¹⁰ He did blame Cyriac for making Perseus left-handed, but wrongly: the illustrator has done this, whether by switching the sword and Medusa's head or by having him face the front and not turn his back on the reader as he does in the scholion Sangermanense.

¹¹ At 357 it turns a genibus into adienibus; cf. Kauffmann, p. 24 on geniali for hiemali in the Vercellensis, confirmed by geniali or gemali in G.

¹² I have noticed such a ligature in P of Germanicus's Aratea (Paris. Lat. 7886, s. ix 3/4, from Corbie) on f. 15r at the foot of the second column, and Bruce Barker-Benfield kindly refers me to another in Lupus's notes on Berlin Phill. 1872 f. 101r. At 216 torquen must lie behind torqueri in AM.

¹³ I have trusted Soubiran's report of HDAM. H² has no authority, and BCKLT should be ignored except where H has lost leaves; V and S should be ignored altogether. I throw in one or two readings that a strict application of the stemma would exclude, mostly where G agrees with H against DAM, and I report G wherever Soubiran prints a reading on the authority of D or AM. Orthographical variants I largely ignore.

¹⁴ McGurk, P., Catalogue of astrological and mythological illuminated manuscripts of the Latin Middle Ages IV (London, 1966), 82 + Pl. IIIa. I have inspected photographs at the Warburg Institute. For evidence of its proximity to G see n. 55.Google Scholar

¹⁵ Ep. 1.304 (to Niccoli from Rome, 27.12.1429): ‘Nam de Frontino et fragmento Arati quod scribis, illi apud me cunt.’ One thinks of two Sicilians interested in manuscripts, Panormita and Aurispa. Neither was in the south at the time, but Panormita had met Poggio at Rome in 1428; see Resta, G., Diz. biog. degli italiani, 7 (Rome, 1965), p. 401.Google Scholar

¹⁶ On the inadequacy of his remarks about the manuscripts see Hall, J. B., Proc. Afr. Class. Assoc. 14 (1978), 44–51. His microfilms are in the Senate House, London University, to which I am indebted for the loan of them.Google Scholar

¹⁷ Verzeichnis astrologischer and mythologischer illustrierter Handschriften des lateinischen Mittelalters in römischen Bibliotheken (Sitzungsber. der Heidelberger Akad. 1915. 6–7), 4.Google Scholar

¹⁸ He owned S, Laur. Strozz, 46, and himself wrote V, Vat. Lat. 3110, probably in the 1380s; see Ullman, B. L., The Humanism of Coluccio Salutati (Padua, 1963), p. 168Google Scholar + Pl. VII, 2, 188–9, and Mare, A. C. de la, The Handwriting of Italian Humanists I, i (Oxford, 1973), p. 41.Google Scholar

¹⁹ An illustration can be seen in Warner, G. F., Descriptive catalogue of illuminates manuscripts in the library of C. W. Dyson Perrins (Oxford, 1920), II, Pl. LXIV, and another in Sotheby's catalogue for 9.12.58, Pl. 39; Dr de la Mare identified the scribe from them, and both show the collator, whose hand resembles Panormita's and Pontano's. I am indebted to the Warburg Institute for access to photographs (not complete) of this manuscript and also of Pierpont Morgan M 389, the Neapolitanus mentioned below (n. 21), and Aberystwyth 735C (see n. 44).Google Scholar

²⁰ Even if Τ existed, it did not omit vv. 70 and 96 as Gain says (p. 8), because U has both; the initial letters of vv. 71 and 97 are nevertheless illuminated, a further sign of contamination. Neither Gain nor anyone else reveals that σ omitted 263 as well as 70 and 96.

²¹ Errors of x include 77 honor, 254 palmis, 266 multo, 352 non om., fr. iv 58 composuit, 74 ingreditur, 94 conspecta signa, 109 ostendunt, 144 ingreditur aurea (in the last four passages I have been able to check only T of the other family); errors of the other family include 75 erit om., 150 huc, 313 rapit, 340 om., 361 piscem, 366 undas, 401 centra, 420 placatam. Gain's stemma below ξ seems wrong in several points. Unless I am much mistaken, Montpellier 452 derives from Madrid Bibl. Nac. 8282, Vat. Lat. 1653 (and why not also 3293?) from the editio princeps, and Berlin Lat. Oct. 149 from the manuscript at Palermo, which together with Vat. Reg. Lat. 1801 derives through an intermediary from British Library Add. 15819. Gain's ∈ does not derive from the source of Vat. Urb. Lat. 1358 and Laur. 89 sup. 43, twins shown by the illustrations in Saxl-Meier, op. cit. (n. 2), liv-lviii, to be in the same hand. I have noticed only the most tenuous evidence (nothing more solid than the subscription ‘non sene trova piu’) for grouping these twins with Madrid 8282 and Vat. Barb. Lat. 77. Dr de la Mare tells me that all four are Florentine, and she has touched on Add. 15819 (‘Florence, c. 1465–75?’) in Cultural Aspects of the Italian Renaissance: Essays in Honour of P. O. Kristeller, ed. C. H. Clough (Manchester, 1976), 185, No. 65. Fonzio's corrections in Add. 15819 recur in the editio princeps, in Eton 88 (no later than s. xv 3/4 and attributed by Dr de la Mare to the Florentine scribe Dominicus Brasichillensis), and in a manuscript unknown to the editors, Naples Naz. XIV D 37, on which see McGurk, op. cit. (n. 14), pp. 62–4 (Dr de la Mare attributes it to G. A. Vespucci); the relationship of these four witnesses puzzles me, except that if the Neapolitanus omits 125–6 vulguspatrum it will have been copied from Add. 15819 after correction. Gain also missed Dyson Perrins 84 (see above); Ambros. D 52 inf., presumably related as in Cicero to Montpellier 452; Siena Com. L VI 26, presumably copied as in Cicero from Montpellier 452; and an edition roughly contemporary with the editio princeps, namely I.G.I. 779.

²² For an illustration see Saxl-Meier (n. 2), p. lii. What remains of the arms fits Fabio Mazzatosta; cf. Maia 27 (1975), 241 n. 51. On Michael Laurentii see CQ 71 (1977), 233 + n. 96.

²³ Winterfeld, op. cit. (n. 6), pp. 395–6. In a work of 1912 that I have not seen, J. t Burnam assigned M to Ripoll in Catalonia, and he has been followed by Vallicrosa, J. Millàs, Assaig d'història de les idees fisiques i matemàtiques a la Catalunya medieval (Barcelona, 1931), pp. 237–40,Google Scholar and Cordoliani, A., Revista de Archivos Bibliotecas y Museos 57 (1951), p. 7Google Scholar and Anuario de Estudios Medievales 3 (1966), pp. 66–7; but half of Burnam's reasons as Millàs Vallicrosa reports them are quite plainly bad, and Cordoliani in his second article not only speaks of M as ‘venant du Mont-Cassin’ but also argues that at least part of it was copied from a Beneventan exemplar. Bordona, J. Dominguez, Manuscritos con pinturas (Madrid, 1933),Google Scholar No. 411, says ‘no parece que el códice sea originario del escritorio de Ripoll, come supone Burnam, sino de Italia’. For plates of the script see the lnventario general de manuscritos de la Biblioteca Nacional I (Madrid, 1953).Google Scholar

²⁴ Boll, F., Sphaera (Leipzig, 1903), p. 445;Google Scholar Saxl-Meier, op. cit. (n. 2), pp. xxxv-xliii. The quotation is from Thomdike, Lynn, Michael Scot (London, 1965), P. 1.Google Scholar

²⁵ Winterfeld, loc. cit. (n. 23). His remarks about Metz, however, make too much of Berlin Phil. 1832, especially since it was written at Laon and only later passed to Metz; see Contreni, J. J., The Cathedral School of Laon from 850 to 950 (Münchener Beiträge zur Mediävistik and RenaissanceForschung 29, 1978), p. 125.Google Scholar

²⁶ On Montecassino and Salerno as the chief repositories of old learning in the kingdom of Sicily see Niese, H., Hist. Zeitschr. 108 (1912), 480–2.Google Scholar

²⁷ On U see Les Manuscrits classiques latins de la Bibliothéque Vaticane I (Paris, 1975), p. 125.Google Scholar

²⁸ V f. 54r ‘Inveni librum metricum et prosaicum cuius titulus erat “T. Claudii Cesaris Arati phenomenia de celi positione et quinque circulis mundi”. Ex hoc libello excerpsi solummodo carmina, sed in alio volumine tam versus quam prosae scripts’ Sabbadini, , S.I.F.C. 7 (1899), 115–18,Google Scholar showed that Tortelli before 1449 consulted a manuscript like SV. The praenomen T. also survives in Aberystwyth 735C (see n. 44), and a lost manuscript catalogued c.1049–1160 at Lobbes (near Liège) had it too; see Dolbeau, F., Recherches Augustiniennes 13 (1978), 33,Google Scholar No. 302. I have remarked elsewhere on the haphazard transmission of titles: CQ 72 (1978), 231 n. 29.Google Scholar

²⁹ On the earliest Latin chartacei see Thompson, E. M., Encyc. Brit., ed. 11, XX (1911), 726,Google Scholar who mentions A without saying more than that it is ‘in an Italian hand of the first half of the 13th century’; he had inspected it for Baehrens (P.L.M. 1.143). It appears to be the earliest chartaceus of a Classical Latin text. For a sample of the script see Catalogue of manuscripts in the British Museum I: the Arundel manuscripts (London, 1840), Pl. III.Google Scholar

³⁰ Burnett, C. S. F., Journal of the Royal Asiatic Soc. 1977, 62–108, with an edition of the preface from A and Bodl. 430; see n. 34 on Michael Scot.Google Scholar

³¹ Lemay, R., Abu Ma'shar and Latin Aristotelianism in the twelfth century (Beirut, 1962), p. 30.Google Scholar

³² Eratosthenis catasterismorum reliquiae (Berlin, 1878), 201–20. His stemma (p. 209) does not fit his own evidence, which shows that BP had a common source below β; corrected accordingly, it matches the corresponding part of Gain's stemma for Germanicus (Robert's 's O).Google Scholar

³³ Gain says nothing about them, but his assertion that ‘only M of the μ manuscripts conforms completely to its stemmatic position’ (p. 8) would be disturbing if it were true. I have no idea what lies behind it.

³⁴ Opp. citt. (n 5).

³⁵ All the verses omitted by μ were already omitted by O, the common source of μ and ν.

³⁶ Ff. 92v-95v ‘Puplii Virgilii Maronis astronomice artis liber’ (a cento, Walther 17030); f. 95v (untitled) Sol duabus unius orbis ultimis partibus—eadem spatia conficiunt (Cic. De nat. deorum 1.87).

³⁷ Winterfeld, op. cit. (n. 6), p. 396 n. 7, contested its authenticity on stemmatic grounds. Gain meets only Breysig's argument (ed.², pp. xxviii-xxix).

³⁸ Martin, op. cit. (n. 5), pp. 38–9, wrongly says that M contains the scholia Basiliensia.

³⁹ The lacunae in χ are easily reconstructed, but not so those in σ, because T omits most of the detached phrases between lacunae (I have not seen this part of Dyson Perrins 84 or Pierpont Morgan M 389). The supplements in U may have come from its secondary source. Supplements also occur in Montpellier 452 (Gain's microfilm begins after Arati genus, but Siena L VI 26 and the ed. Ven. 1488 both have them); to judge from its readings, e.g. Gecraustius where μ read Grecaustius, it may have drawn on a manuscript of the scholia Sangermanensia rather than Germanicus, but it certainly drew on another descendant of μ for the end of the scholion on Hydra (181.14 tres—18 quadraginta tres). Laur. 89 sup. 43 has the same supplements, entered later by the scribe himself.

⁴⁰ Perhaps, therefore, the illustration of Aratus and the Muse, which follows Arati genus in M, preceded it on a separate page in σ or its exemplar and later came adrift. Otherwise it would be hard to see why σ should omit it.

⁴¹ Being the only manuscript to contain the scholia Strozziana, μ alone depicts a female Centaur after the passage cited by Breysig in the apparatus on 112.3. The source of this illustration has not been traced; see Boll, op. cit. (n. 24), pp. 445–6, and Sod-Meier (n. 2), p. xxxviii. On the illustrations in Germanicus and other astronomical texts see Thiele, G., Antike Himmelsbilder (Berlin, 1898), Ch. IV; A. W. Byvanck, Med. Kon. Ned. Akad. Wet. N.R. 12 (1949), 169–235 (to be used with care); Saxl-Meier, op. cit., pp. xiii-lix; Martin, op. cit. (n. 5), pp. 48–51; McGurk, op. cit. (n. 14), pp. xiii-xxv, and Nat. Lib. of Wales Journal 18 (1973–4), 197–216.Google Scholar

⁴² Winterfeld, op. cit. (n. 6), pp. 393–5.

⁴³ Bischoff, , Mittelalterliche Studien I (Stuttgart, 1966), p. 59.Google Scholar

⁴⁴ McGurk discusses Ab, which editors have not used, in the article of 1973–4 cited above (n. 41). He does not say where it puts the illustrations, but I have looked at photographs in the Warburg Institute.

⁴⁵ See D.A. 28 (1968), 3656A; the Bodleian has an authorized copy (2954 e 1). The Teubner of B. Bunte (Leipzig, 1875), founded on three manuscripts from German libraries when over 50 exist, has survived repeated exposure of its inadequacy and inaccuracy and now reappears in Italian dress (Pisa, 1976). Sister Fitzgerald's edition is preceded by a list of 61 manuscripts, accompanied by a collation of 28, and followed by a discussion of these 28.

⁴⁶ She somehow conflates the latter with Vindob. 2269 in her description (p. 27).

⁴⁷ Errors of a that appear in no manuscript reported by Sister Fitzgerald include 105.4 [ad] antarcticum autem circulum tangit, 105.9 pennam sinistram quae ultra, 106.9 difficilis om., 110.19 auribus, 112.5 corpore et, 12 permanentem semper, 24 <necesse est> effugere, 113.12 posse <constat>, 20 occasum <non>, 9 consistat quaelibet autem navis.

⁴⁸ I do not know why Saxl, loc. cit. (n. 17), says that on the evidence of U ‘es ist …sehr wahrscheinlich, dass Poggios Hs. in karolingischer (spätestens ottonischer) Zeit geschrieben war’, or why Byvanck, op. cit. (n. 41), p. 7, adds that ‘de humanisten beschouwden het als een codex, die uit de klassieke Oudheid stamde’. As Ullman, B. L. says in Studi in onore di Luigi Castiglioni (Florence, 1960), II, p. 1037, ‘it is a universal assumption, inherited from the period of romanticism, that every lost manuscript must have been both old and valuable.’Google Scholar

⁴⁹ For a table of the poems as they appear in M, S, and Cava 3, see Cordoliani, A., Anuario de Estudios Medievales 3 (1966), pp. 74–6; his derivation of S from M flies in the face of a simple fact exhibited on p. 75. I have consulted the Liber introductorius in MS Bodl. 266, where ff. 93r–94r present Nos. 2–6, 9, and 13, of the 19 pieces tabulated by Cordoliani.Google Scholar

⁵⁰ Prete, S, Ausonius (Teubner, Leipzig, 1978), 104.Google Scholar

⁵¹ According to Prete, K³ reads unde december amat to genialis hiemps, but Tobin, N. W., The text of the Eclogae of Decimus Magnus Ausonius (diss. Fordham, 1967), p. 35, explains that it fuses Primus Romanas with Dira patet lani (Anth. 394) and puts the last two lines in the wrong order; unde December is actually the last line of Dira patet lani, and the line that belongs to Primus Romanas is the usual interpolation, imbrifer ast mensis tumque december adest. On the other hand, Tobin says that A⁴ omits the line (p. 36), while Prete cites a variant from it. I believe Tobin about K³ but do not know whom to believe about A⁴.Google Scholar

⁵² Cf. Cordoliani, op. cit. (n. 49), pp. 67, 74 + n. 52.

⁵³ T and its relatives put Capricorn at the appropriate point in the scholion, but the other manuscripts all agree against them.

⁵⁴ McGurk, op. cit. (n. 44), regards the text of Ab as hybrid and its illustrations as significantly different from those in other manuscripts of Germanicus; but his evidence on the first point will not bear inspection, and even if contamination accounts for certain similarities between its illustrations and those in the pseudo-Bedan catalogue of stars, the presence both of Aratus and the Muse and of Jupiter and the eagle shows that they belong fundamentally to Germanicus.

⁵⁵ Martin (n. 5), p. 44, gives the fullest list of manuscripts. Add G of Cicero (see part I) and Siena Corn. L IV 25 (n. 14). From photographs at the Warburg Institute I have collated G and the Senensis where they overlap, namely in pp. 109–204.2 and 210.4–238.5 of Maass, op. cit. (n. 5), and in the same passages Maass's DKS^II and such parts as were visible of S^I, the Sangermanensis, and Paris Nouv. Acq. Lat. 1614 (N in what follows). G and the Senensis share e.g. 190.2 ut circo, 192.4 <et> Bacchus, 194.6 <et> in capite, 213.6 extendit, 221.1–2 summa for sub uno, 6 earn, 221.1 proiecta, 231.2 vermes for nervi. In several places they stand with K and the Sangermanensis against DS^IS^II, which read e.g. 190.2 [a love] (so too N), 197.3 [ideo], 199.1 unam for singulas, 201.5 [earn], 203.3 [signis], 212.2 toniadae, 223.2 delton, 224.1 asingulis for angulis (so too N), 235.3 [ut], 238.2 quo. DS^IS^II are clearly wrong at 190.2, 199.1, 212.2, 223.2, 224.1, 238.2, and N presumably agrees with them in all these places; whether the rest also form a group is not clear. Before S^II was corrected from outside its group, it closely resembled S^I, but it does not seem to have been copied from it: 196.5 [labio] S^I. These remarks are compatible with Manitius's stemma for D, K, and the Sangermanensis, op. cit. (n. 5), p. 308 n. 1, but conflict with Martin's assertion, op. cit., p. 44, that N resembles the Sangermanensis, and with the assertion I have read somewhere that S^II was copied from S^I. Maass did not see that the lists of stars given by D at the end of the sections on Lyra, Cycnus, Aquarius, and Capricomus, were interpolated from the pseudo-Sedan catalogue of stars.

⁵⁶ Thiele, op. cit. (n. 41), p. 82.

⁵⁷ Catalogue général des manuscrits des bibliothèques publiques des départements IV (Paris, 1872), p. 687.Google Scholar

⁵⁸ Dahms, R., Jahrb. für class. Phil. 99 (1869), 270–1.Google Scholar

⁵⁹ Cf. Breysig² xv, ignored by the latest champion of the derivation, Meyier, K. A. de, Codices Vossiani Latini II: codices in quarto (Leyden, 1975), p. 187.Google Scholar

⁶⁰ For a specimen of both scripts see Thiele, op. cit. (n. 41), p. 78.

⁶¹ I examined L in August 1979.

⁶² Gain, p. 8. Le Boeuffle, p. xlvi, says that L and E differ in 110 places, a figure useless until it is broken down; a quick look through his apparatus reveals at most 55 errors of L, many of them trivial but some of more weight, e.g. 276, 335, 367, 421, 441.

⁶³ Le Boeuffle, p. xxvi, Gain, p. 124.

⁶⁴ Thiele, op. cit. (n. 41), pp. 79–80. Gain nowhere refers to Thiele, and his own description of L, pp. 5–6, is less helpful.

⁶⁵ pp. 5–8.

⁶⁶ Kristeller, P., her Italicum II (Leyden, 1967), 46, cited in CQ 71 (1977), 223 n. 97Google Scholar

⁶⁷ Boutemy, A., Annales de la Soc. Arch. de Namur 41 (1934), 43–85Google Scholar and Mélanges Félix Rousseau (Brussels, 1958), pp. 111–20Google Scholar

⁶⁸ Philologus 66 (1907), 122–3.Google Scholar

⁶⁹ Dolbeau, loc. cit. (n. 28).

⁷⁰ Becker, G., Catalogi bibliothecarum antiqui (Bonn, 1885), 68, No. 210 (pp. 152–3).Google Scholar

⁷¹ His collation has been used only in Bentley's copy, but the original survives in Leyden 755 H 15 (ed. Lugd. 1566). It mentions neither the identity nor the age of the manuscript, but Jacobus Gronoviu; was somehow able to inform Bentley of both; perhaps a note has come adrift in rebinding.

⁷² People repeat Sabbadini's inference in Le scoperte dei codici latini e greci ne' secoli XIV e XV, I (Florence, 1905), p. 170Google Scholar that Leone's manuscript was ‘in maiuscolo o comunque anteriore al sec. IX’, but surely Politian just meant that he had seen no older manuscripts of Manilius. Transcripts of other texts that it contained survive in Munich; see Pierro, C. di, G.S.L.I. 55 (1910), 9.Google Scholar The only reading attested in Manilius, 5.126 ut fidum nerea dii genuere syboeten, for which see Angelo Poliziano, Miscellaneorum centuria secunda, ed. Branca, V. and Stocchi, M. Pastore (Florence, 1972), IV 34, recurs (except for the triviality syboten) in Gronovius's collation of the Venetus; other manuscripts have et fidunt followed by something like nerciadu.Google Scholar

⁷³ Weigle, F., Die Briefsammlung Gerberts von Reims (Weimar, 1966), Ep. 8, 130. If Bonincontrius really received an old Casinese fragment from Panormita, it should have made more of a difference to his edition (Rome, 1484).Google Scholar

⁷⁴ See however CQ 71 (1977), 223.Google Scholar

⁷⁵ Garrod, , Manili Astronomicon liber II (Oxford, 1911), p. xxv;Google ScholarGoold, , Phoenix 13 (1959), 96, ed. (1977), cvi.Google Scholar

⁷⁶ It is absent from ‘Latin classics’, but I was looking through ‘Astronomy’. It now appears in the alphabetical catalogue on cards.

⁷⁷ The note was published by Lehmann, P., Sitzungsber. der bayer. Akad. (1930), 2, p. 8,Google Scholar who said of the manuscript ‘wohl aus Amorbach’. In Studien and Mitt. zur Gesch. des Benediktinerordens 48 (1930), 288Google Scholar = Erforschung des Mittelalters III (Stuttgart, 1960), p. 97Google Scholar he retracted the attribution for a reason that seems to be refuted by the table of contents. Signed manuscripts of Sifridus Schlunt that Lehmann mentions include one written at Amorbach in 1448 and one at Gotthardsberg in 1466; British Library Add. 18972, Colophons V (1979), 17074, was written at Amorbach in 1451. The watermark in A is Briquet 12995: transalpine, mainly German, c. 1430–49.Google Scholar

⁷⁸ Cappelli, A., Dizionario di abbreviature latine ed italiane (Milan,⁴ 1949), p. 154,Google Scholar gives ‘xv p.’ (= s. xv^I), but Loew, E. A., The Beneventan script (Oxford, 1914), p. 181,Google Scholar says that in Beneventan ‘gs(saec. xi) is the normal form’, and Bains, Doris, A supplement to Notae Latinae (Cambridge, 1936), pp. 17–18, cites examples from Rome Naz. Sess. 44 and 45 (s. xi¹, from Nonantola) and Lambeth Palace 431 (s. xi, from Llanthony).Google Scholar

⁷⁹ Lehmann, P., ‘Die mittelalterliche Dombibliothek zu Speyer’, Sitzungsber. der bayer. Akad. 1934. 4, pp. 23–7.Google Scholar

⁸⁰ See part II.

⁸¹ Dolbeau, op. cit. (n. 28), p. 33, No. 303 and ibid., 14 (1979), No. 303.

⁸² Meier, G., Jahrb. für schweiz. Gesch. 10 (1885), 112–14.Google ScholarClark, J. M., The Abbey of St as a Centre of Literatury and Art (Cambridge, 1926), p. –,Google Scholar says that about AD 1000 the Irish communities of St Gallen and Liége were in close contact; Notger, bishop of Liege from 972 to 1008, had been ‘prepositus monasterii Sancti Galli’ according to one source, but Kurth, G., Notger de Liége et la civilisation au x^e siécle (Paris, 1905), pp. 35–7, is surely right to suspect confusion with some other Notger.Google Scholar

⁸³ For help in the preparation of this article I should like to thank Bruce BarkerBenfield, Charles Burnett, Albinia de la Mare, Carlotta Griffiths, David McKie, Nigel Palmer, and Richard Rouse.