¹ For a thorough presentation of eighteenth and nineteenth-century controversy on the Frankish aspects of this subject, see Blancard, Louis, ‘La question Gondovald,’ Mém . Acad. Marseille (1884–5) 409–42. Amédée Gasquet, L'empire byzantin et la monarchie franque (Paris 1888) 183–204, devotes a chapter to these years but seriously overestimates Byzantine designs on Gaul. Georges Reverdy's ‘Les relations de Childebert II avec Byzance,’ Revue historique 114 (1913) 61–86. is primarily concerned with the years following 584. In his recent book, Byzance avant l'Islam II 1: Byzance et les Francs (Paris 1956), Paul Goubert has discussed the entire Gundovald affair; a preliminary study, ‘L'aventure de Gondovald et les monnaies franques de l'empereur Maurice,’ Échos d'Orient 39 (1941–2) 414–43, contains several points not repeated in the book. That I am seldom able to agree with Fr. Goubert's interpretations will become apparent; for one thing, he has not seized the intimate connection between Hermenegild's enterprise and Gundovald's, although he wrote at length on Hermenegild in his article ‘Byzance et l'Espagne wisigothique (554–711),’ Études byzantines 2 (1944) 5–78. As regards Frankish internal affairs, so crucial in these events, none of the foregoing authors, except Goubert, made use of Godefroid Kurth's great monograph, ‘La reine Brunehaut,’ Études franques (Paris 1919) I 265–356. — In addition to the conventional sigla, CSHB will be used for Corpus scriptorum historiae Byzantinae (Bonn 1828ff.), and HF for Gregory of Tours, Historiarum libri X (edd. Krusch, Bruno and Levison, Wilhelm [K-L]: MGH Scriptores rerum Merovingicarum 1 part 1 [2nd ed. Hannover, 1937–51]; tr. Dalton, O. M. [Oxford 1927] 2).Google Scholar

I am very much indebted in the preparation of this article for the assistance and invaluable criticism of Professor Robert L. Wolff of Harvard University. My thanks also go to Dr. Giles Constable, of the State University of Iowa, whose doubts were the genesis of this study and whose comments at a later stage were of no less help; I am likewise grateful to Dr. M Robert, O'Clair and Mr, Carl B, Schmitt, both of Harvard.

² Protector, Menander, Excerpta de legationibus barbarum ad Romanos 5.11–14 (edd. Bekker, A. and Niebuhr, B. G. CSHB [Bonn 1829] 286–90, 303–10); to my knowledge, the best modern account is Ljudmil Hauptmann's ‘Les rapports des Byzantins avec les Slaves et les Avares pendant la seconde moitié du vi ^e siècle’ (tr. Backvis, C. ), Byzantion 4 (1927–8) 151–2; cf. also Thomas Hodgkin, Italy and Her Invaders (2nd ed. Oxford 1916) V 138–9; Goubert, Paul, Byzance avant l'Islam I: Byzance et l'Orient sous les successeurs de Justinien: L'empereur Maurice (Paris 1951) 52.Google Scholar

³ Hodgkin V 162; for a precise listing, see Schmidt, L., Geschichte der deutschen Stämme bis zum Ausgang der Völkerwanderung I: Die Ostgermanen (2nd ed. Munich 1934) 592.Google Scholar

⁴ Goubert, , Byzance et l'Orient 55.Google Scholar

⁵ See below 80–81 and nn. 30, 33, 35. Google Scholar

⁶ This religious motive is fully expressed in Procopius, History of the Wars 5.5: ‘ “… it is proper that you should join with us in waging this war, which is rendered yours as well as ours … by the orthodox faith, which rejects the opinion of the Arians …” ’ (tr. Dewing, H. B., Loeb Classical Library, Procopius 3 [Cambridge, Mass. 1953] 45).Google Scholar

⁷ Ibid. : ‘… making a gift of money to them, he [Justinian] agreed to give more as soon as they should take an active part. And they with all zeal promised to fight in alliance with him,’Google Scholar

⁸ Procopius 7.23 (4.437–9 Dewing). Frankish duplicity is a constant theme in Procopius 5.13; 6.12, 28 (3.137–41, 395–7; 4.115–21 Dewing); Ernst Stein, Histoire du Bas-Empire (Paris-Brussels-Amsterdam 1949) 354–5, 365: some lesser instances. Google Scholar

⁹ Procopius 6.25; 8.24, 26 (4.85–93; 5.305–13, 333–5 Dewing); HF 3.32 (128 K-L; 2.108 Dalton): Gregory is very badly informed; Georg Löhlein, Die Alpen- und Italienpolitik der Merowinger in VI. Jahrhundert (Erlanger Abhandlungen zur mittleren und neuen Geschichte, edd. Schniedler, B. and Brandt, O. 17; Erlangen 1932) 4–43: very detailed, but considered somewhat exaggerated by Stein, op. cit. 361 n. 2.Google Scholar

^9a Myrinaeus, Agathias, Historiae 1.6–7, 2.1–10 (ed. Niebuhr, B. G., CSHB [Bonn 1828] 26–30, 63–88); the minor sources are cited in Löhlein 43–8, and Stein 605–10.Google Scholar

¹⁰ Menander 2 (345–6 Bekker-Niebuhr); Theophanes, Chronographia under 6055 (ed. Classen, Johannes. CSHB [Bonn 1839] 1.367; ed. de Boor, C. [Leipzig 1883] 1.237); Paul the Deacon, Historia Langobardorum 2.2 (ed. Bethmann, L. and Waitz, G. MGH Scriptores rerum Langobardicarum et Italicarum saec. VI-IX [Hannover 1878] 73; tr. Foulke, W. D. [Philadelphia 1907] 55); Liber pontificalis (ed. Duchesne I, Louis [Paris, 1886] 305); Agnellus, Liber pontificalis ecclesiae Ravennatis 90 (ed. Holder-Egger, O. MGH Script. rer. Lang. 336); Stein 610–11; id., Studien zur Geschichte des byzantinischen Reiches vornehmlich unter den Kaisern Justinus II. u. Tiberius Constantinus (Stuttgart 1919) 15, 34 n. 16; Hartmann, L., Geschichte Italiens im Mittelalter I (Leipzig 1897) 348, 404 n. 1; Löhlein, Italienpolitik 49–50; Robert Holtzmann, ‘Die Italienpolitik der Merowinger und des Königs Peppin,’ Das Reich: Idee und Gestalt. Festschrift für Johannes Haller, edd. Dannenbauer, H. and Ernst, F. (Stuttgart 1940) 105.Google Scholar

¹¹ Epistolae Austrasicae 13 (ed. Gundlach, Wilhelm, MGH Epistolae 3 [Berlin 1892] 127–8). The superscription reads: ‘Domino suo, Chamingo duce, Gogo,’ and the letter shows that Amingus was Gogo's patron at court; this connection is exceptionally interesting in the light of Gogo's importance at the Austrasian court during and after Sigibert's reign, for which, see infra. ^11a Agathias 1.5 (23–6 Niebuhr); Hodgkin V 14–5.Google Scholar

¹² Agathias 1.2 (17 Niebuhr). Aside from their naïve tone the descriptive chapters on the Franks contain outrageous errors; but it is suggested that these passages should be considered in the same light as certain fanciful sections of Tacitus’ Germania: Extraits des auteurs grecs concernant la géographie et l'histoire des Gaules 5 (ed. Cougny, E., Société de l'histoire de France; Paris 1886) 415 n. 1.Google Scholar

¹³ Although this attitude might be reflected in Agathias’ good opinion, actual proof that it was held at court is to be found in the transactions about to be described. Google Scholar

¹⁴ HF 4.40 (172 K-L; 2.149 Dalton; cited n. 1 supra); Stein, Studien 16, 34f. n. 18. In the latter passage, Stein decisively rejects the contention of Carrière, A., ‘Sur un chapitre de Grégoire de Tours relatif à l'histoire d'Orient,’ Annuaire de l'École pratique des hautes études (1898) 5–23, that this embassy left in 574. On Sigibert and the Avars, see: HF 4.23, 29 (155, 161–2 K-L; 2.134, 138–9 Dalton); Menander 10 (302–3 Bekker-Niebuhr); cf. an interesting comment in Hauptmann, op. cit. (n. 2 supra) 150.Google Scholar

¹⁵ HF 4.40 and Stein as in n. 14. Google Scholar

¹⁶ ‘ Incipit dicta Gogone ad Grasulfo de nomen regis’: Epistolae Austrasicae 48 (152–3 Gundlach; n. 11 supra). The grammar of the heading is only a forestaste of worse to come. A free translation is to be found in Hodgkin VI 46–7.Google Scholar

¹⁷ The letter begins: ‘Rem necessariam et valde partibus oportunam celsitudo vestra per Billulfo parenti vestro innotuit, quam oportet fixo ordine pro resecanda contumacia intestantium celeriter confirmare. Et licet piissimus imperator revertentibus laegatariis nostris sacris principalibus indicavit, legationem suam confestim velle ipsorum e vestigio ad nos dirigere, quam diebus singulis et venturam credimus et votis ambientibus excipere exoptamus; …’ In the balance of the message, Grasulf is offered alternative courses of action to be pursued (ibid.); there is also an allusion to winter (153 line 2). Google Scholar

¹⁸ HF 6.1 (265 K-L; 2.233 Dalton) marks Gogo's death. As this is the only verifiable information, Gundlach (MGH Epistolae 3.152) simply marks the letter: ‘ante 581?’ This happily does away with the later date offered by Carlo Troya, Codice diplomatico Langobardo I (Naples 1852) 105, 110 f. and Hodgkin VI 45 n. 3. F. Dölger, Regesten der Kaiserurkunden des oströmischen Reiches 1 (Munich and Berlin 1924) 10 no. 76, does not accept this limit, saying that the Gogo of the superscription need not be identical with the known Gogo, and that the heading may simply be a mistake of the assembler of the collection. In his view, the embassy which the emperor promised to send is probably the one which brought Childebert 50,000 solidi c. 583 (see 110–11 and nn. 170, 174 infra). However, since there is no certainty that there was a Byzantine embassy to Childebert c. 583 or indeed before spring 585, it seems to me that dating this letter c. 571–572 better accounts for the known circumstances. Google Scholar

¹⁹ Stein, , Studien 108.Google Scholar

²⁰ Reverdy, , op. cit. (n. 1 supra) 64: ‘Le gouverneur de Childebert, Gogon, écrivit au nom de son maître à Grasulf …’ Löhlein, Italienpolitik 64–5, and Goubert, Byzance et les Francs 197, similarly consider Gogo only as Childebert's nutritor. The only other Merovingian letter written during a king's minority (Epistolae Austrasicae 18; 131 Gundlach) does not have a heading similar to that of the letter in question. It appears more likely to me that a strong, confident and active king like Sigibert would have delegated one of his high officials to write this letter, than that a regent should have omitted the young king's name; we are still far from the Carolingian mayors of the palace. However, any arguments from the heading are at best tentative since the extant document is a chancellery copy rather than the actual letter to Grasulf; moreover, it is the last letter of the collection, and the heading might, from the later copyist's point of view, be the easiest part to abbreviate.Google Scholar

²¹ Fortunatus, Venantius, Opera poetica 7.1.41–2: ‘nuper ab Hispanis per multa pericula terris / egregio regi gaudia summa vehis.’ (ed. Frederick Leo: MGH Auctores antiquissimi 4 part 1 [Berlin 1881] 154). The poem is adressed ‘Ad Gogonem’; on the date, see Meyer, W., Der Gelegenheitsdichter Venantius Fortunatus (Abh. Akad. Göttingen N.F. 4.5 [1901] 14).Google Scholar

²² The next Austrasian known to have gone to Constantinople is Duke Guntram Boso in 581: infra 96 and n. 97. No Byzantine came to Austrasia before 585: infra 116 and n. 194. Google Scholar

²³ Hodgkin VI 37, 45, argues that they were brothers, on the basis of Epistolae Austrasicae 41, a letter from Exarch Romanus to Childebert II written perhaps in 590: ‘We … marched into the province of Istria against our enemy Grasulf. His son, the magnificent Duke, Gisulf, wishing to show himself a better man than his father, came with his nobles and his entire army, and submitted himself to the Holy Republic’ (147 Gundlach; tr. in Hodgkin V 273). Hogdkin concludes that Gisulf I was succeeded in Friuli (of which Istria was a part) by his brother Grasulf, who because of age had by 590 associated his son Gisulf II to himself in the dukedom. Goubert, Byzance et les Francs 197, has joined Hodgkin in this opinion. Google Scholar

²⁴ Paul the Deacon 2.9 (65–6 Foulke; n. 10 supra): ‘… [Alboin] determined … to put over the city of Forum Julii and over its whole district, his nephew Gisulf, who was his master of the horse … a man suitable in every way. This Gisulf announced that he would not first undertake the government of this city and people unless Alboin would give him … the families or stocks of the Langobards that he himself wished to choose. And this was done, and with the approval of the king he took to dwell with him the chief families of the Langobards he had desired … He asked also from the king for herds of high-bred mares, and in this also he was heeded by the liberality of his chief.’ Google Scholar

²⁵ Hodgkin VI 45, considers that the expression ‘vestra celsitudo’ in Gogo's letter implies that Grasulf had already succeeded Gisulf in the duchy of Friuli, which would date the letter in the late 570's. However, if this were so, why would Gogo not have addressed Grasulf by his proper title? Calling him ‘vestra celsitudo’ might only be a way to flatter his vanity. Reverdy (n. 1 supra) 63–64, also believes that Grasulf had already succeeded to the duchy, but only because he dates the letter from Gogo's regency, that is, after 575 (loc. cit. 64 n. 1). Google Scholar

²⁶ Both of the lost documents are mentioned in the extant letter: MGH Epistolae 3.152 lines 23–4 and 35–6. Google Scholar

²⁷ This possibility is strongly suggested by Grasulf's later history (n. 23 supra). Google Scholar

²⁸ Stein, , Studien 16–7, suggestes that the Empire may have had a hand in the murder of Alboin in 572. He considers it a sign of Byzantine involvement that Rosamund and her co-conspirators were welcomed to Ravenna by the Byzantine prefect Longinus. However, I cannot see in this action more than the perpetual Byzantine care to welcome dissident barbarians. The tragic story of Rosamund and her motives for murdering Alboin are well known and leave no room for imperial intrigue. Goubert, Byzance et l'Orient 52, echoes Stein and adds, without adducing evidence, that Byzantium had a hand in fomenting the murder of King Cleph in 574.Google Scholar

²⁹ Goubert, , Byzance et l'Orient 55–6.Google Scholar

³⁰ Stein, , Studien 104; Hartmann, op. cit. (n. 10 supra) II 1 (Leipzig 1900) 46.Google Scholar

³¹ Hartmann, , loc. cit. 38ff., 52 n. 5; Hodgkin V 182ff.Google Scholar

³² Diehl, Charles, Études sur l'administration byzantine dans l'Exarchat de Ravenne (Paris 1888) 17–18.Google Scholar

³³ John of Biclar, Chronica a. DLXVII-DXC under 576 (ed. Mommsen, Th.: MGH Auct. ant. 11 [Berlin 1894] 214); Agnellus 51 (313 Holder-Egger; n. 10 supra); Stein, Studien 103–4; Hartmann II.1.47; Hodgkin V 194–5. Löhlein, Italienpolitik 63–4, argues that the expedition of Chramnichis, known only through Paul the Deacon, Historia Langobardorum 3.9 (97 Bethmann-Waitz; Foulke 102f.), may have been a result of planned Austrasian cooperation with Baduarius’ expedition.Google Scholar

³⁴ Menander 25 (327–8 Bekker-Niebuhr). Google Scholar

³⁵ Menander 29 (331–2 Bekker-Niebuhr). This may well have been the embassy in which the later Gregory the Great came to Constantinople as papal apocrisiary. Both these translations are from the Latin. Google Scholar

³⁶ Paul the Deacon 3.18–9 (101–2 Bethmann-Waitz; 118–9 Foulke). Even this episode is probably not related to the embassy of 579, since it happens some five or six years later. Google Scholar

³⁷ Epistolae aevi Merowingici collectae no. 9 (ed. Gundlach, W.: MGH Epistolae 3.448–9). On Aunacharius (who is called Aunarius in the superscription), cf. MGH Script. rerum Merov. 1 part 2 (Hannover 1885) 653 n. 1.Google Scholar

³⁸ The translation, with a slight correction, from Mann, H. K., ‘Pelagius II,’ Catholic Encyclopedia (N. Y. 1911) 11.603.Google Scholar

³⁹ The names Burgundy, Austrasia, and Neustria are varyingly anachronistic; of the three, only Neustria does not appear at all in Gregory of Tours. So far as this study is concerned, these names are simply labels for the kingdoms of the monarchs specified. Similarly, a ‘Burgundian’ is a subject of King Gunthram, an ‘Austrasian,’ a subject of King Childebert, and so forth. The territorial shifts are greatly clarified by the maps in Auguste Longnon's Géographie de la Gaule au VI ^e siècle (Paris 1878). Google Scholar

⁴⁰ Ma ius of Avenches, Chronica under 569 and 574 (ed. Mommsen, Th.: MGH Auct. ant. 11 [n. 33 supra] 238f.); HF 4.42, 44 (174–6, 178–80 K-L; 2.150–1, 153–4 Dalton); Hodgkin V 216–23. There is no point in citing Paul the Deacon, who only copies Gregory in this case.Google Scholar

⁴¹ Ibid. For the date of the first attack, see Stein, Ernst, ‘Post-consulat et αὐτоϰϱατоϱία,’ Annuaire de l'Institut de philologie et d'histoire orientales 2 (Brussels 1934) 887.Google Scholar

⁴² Troya, , op. cit. (n. 18 supra) nos. 19–21 (I 71–83); Pseudo-Fredegar, Chronicle 45 (ed. Monod, Gabriel, Études critiques sur les sources de l'histoire mérovingienne II [Paris 1885] 137f.); Hartmann II. 1.81 n. 4. Löhlein, Italienpolitik 56–9, has an excellent discussion of this cession; he shows that Susa was an isolated Byzantine outpost, which was probably given up freely to King Gunthram's forces.Google Scholar

⁴³ HF 5.46 (256 K-L; 2.219 Dalton): ‘Transobad … was the most persistent candidate, trusting to the fact that he had placed his son in the household of Gogo, then governor of the king.’ Also HF 6.1 (265–6 K-L; 2.233 Dalton). The corroborative evidence offered by Fortunatus and the Austrasian letters (supra 6–7 and nn. 11, 16, 21) is valuable for confirming Gogo's importance, to which Gregory hardly does justice in his brief references. Even so, Gogo's role can only be appreciated in terms of his policy, the initiative for which Gregory assigns to the child Childebert. Cf. Georg Waitz, Deutsche Verfassungsgeschichte II 2 (3rd ed. Berlin 1882) 107 with n. 1. Google Scholar

⁴⁴ Brunhild's support of Gogo is, in my opinion, shown by comparing HF 5.14 (213 K-L; 2.184 Dalton) with HF 6.4 (267–8 K-L; 2.234–5 Dalton). Kurth, loc. cit. (n. 1 supra) 282f., largely because he believes HF 5.14 to show Gogo deliberately frustrating Brunhild, concludes that Gogo was a moderate notable rather than a devoted royal servant; I cannot agree. Google Scholar

⁴⁵ HF 5.13, 44 (207, 252–4 K-L; 2.179, 217f. Dalton). Google Scholar

⁴⁶ HF 5.13 (loc. cit) Google Scholar

⁴⁷ HF 5.17 (216 K-L; 2.185–6 Dalton). Google Scholar

⁴⁸ HF 8.30, 45; 9.16, 31 (393–4, 411, 430–1, 450 K-L; 2.350–3, 365–6, 385, 401–2 Dalton). Gunthram's enmity toward Visigothic Spain is evident in the later part of Gregory's History; civil disturbances and the Lombards probably explain why there is no record of earlier attempts on his part to seize Septimania. Yet it might be argued that Visigothic fears of Gunthram form the political motive for the Spanish-Merovingian marriages of 566 and 568, HF 4.27–8 (160–1 K-L; 2.137–8 Dalton), and that King Sigibert's seemingly motiveless war for Arles, HF 4.30 (162–3 K-L; 2.139–40 Dalton), was a deliberate, Spanish-sponsored attempt to cut Gunthram off from his only access to Septimania. Google Scholar

⁴⁹ of Biclar, John, Chronica under 579 (215 Mommsen; n. 33 supra).Google Scholar

⁵⁰ See 75–6 and nn. 7, 8 supra. Google Scholar

⁵¹ HF 4.38 (170 K-L; 2.147 Dalton): ‘[Leuvigild already had] two sons by his first wife, one of whom was betrothed to [Ingund] the daughter of Sigibert, the other to [Ringuth] the daughter of Hilperic [i.e. Chilperic].’ It is difficult to say whether Gregory is referring to an actual family compact or is writing with the benefit of hindsight. Google Scholar

⁵² Reverdy (n. 1 supra) 63; Hartmann II. 1 (n. 30 supra) 61.Google Scholar

⁶³ HF 6.2 (266–7 K-L; 2.233–4 Dalton); see infra 92 and n. 82. Google Scholar

⁶⁴ The momentous fact that King Gunthram's great general Mummolus fled the Burgundian court in 581, for reasons not specified, leads me to interpret as signs of unrest a number of fragments not treated as related by Gregory of Tours. Aside from Mummolus, the only known actors are the brother bishops, Salonius and Sagittarius, who, to the Church's horror, had taken an active part with Mummolus in the campaign of 571, HF 4.42 (175 K-L; 2. 151 Dalton), and probably in later ones. In 576 Mummolus still served Gunthram faithfully: HF 5.13 (207 K-L; 2.179 Dalton). But then, Salonius and Sagittarius, clearly a pair of bad characters, were summoned before Gunthram for a hearing about their recent crimes; Sagittarius answered by contesting the legitimacy of Gunthram's sons; this brought the brothers prompt imprisonment, but through the intercession of Burgundian magnates and bishops, they were soon released: HF 5.20 (227–9 K-L; 2.195–7 Dalton). In 577 Gunthram issued an ultimatum to Chilperic, but Mummolus did not enforce the threat after Chilperic had rejected the ultimatum, HF 5.17 (216 K-L; 2.186 Dalton); it is possible, however, that Gunthram gave no orders to that effect. In 579 Salonius and Sagittarius were stripped of their sees by a Church council, but only after Gunthram had especially intervened to add to the existing charges of murder and adultery the further accusations of lése-majesté (rei maiestatis … proditores) and treason: HF 5.27 (233 K-L; 2.200 Dalton). After the party of Austrasian notables seized power at Metz in 581, Mummolus fled to Austrasian Avignon HF 6.1 (266 K-L; 2.233 Dalton), and in the Gundovald rising of 584–85, Sagittarius reappeared to join Mummolus in the attempt to establish a pretender, HF 7.28 (346 K-L; 2. 305 Dalton). Google Scholar

⁵⁵ HF 5.17 (215–6 K-L; 2.185–6 Dalton): ‘He [Gunthram] himself lost two sons, destroyed by sudden disease … Thereafter King Gunthram sent envoys to his nephew Childebert, seeking peace, and praying that they might meet … [Gunthram:] “Let a single shield protect us and a single spear defend I If I should yet have sons, I will none the less regard thee as one of them….” ’ Note the extremely favorable nature of the last provision: even if Gunthram had new heirs, Childebert would still share in the partition of Burgundy after Gunthram's death. Google Scholar

⁵⁶ This party's formation can be assumed from its later history, which spans the entire decade of the 580's after the overthrow of Gogo in 581. Egidius’ ties with Chilperic are revealed at his trial in 590, HF 10.19 (510–2 K-L; 2.455–7 Dalton), which marks the end of an era. Cf. Kurth (n. 1 supra) 283; Walther Schultze, Das merowingische Frankenreich (Deutsche Geschichte von der Urzeit bis zu den Karolingern 2; Stuttgart 1896) 152–3, out-lines the nature of aristocratic opposition. Google Scholar

⁵⁷ HF 4.51, 28 (188, 161 K-L; 2.160, 138 Dalton). Google Scholar

⁵⁸ HF 5.34 (239–41 K-L; 2.205–6 Dalton). Google Scholar

⁵⁹ HF 5.39 (245–7 K-L; 2.210–2 Dalton). Google Scholar

⁶⁰ Venerable but none the less modern tradition places Brunhild's marriage to Sigibert in 566, or with equal certainty in 567. Consequently, Ingund could at the earliest be born in 567. Pending further study of this exceptionally complicated early chronology, an earlier date might well be assumed for Brunhild's marriage, which would in turn make Ingund somewhat older than twelve in 579. Google Scholar

⁶¹ HF 5.38 (244 K-L; 2.209 Dalton). Through an earlier marriage, this woman, strangely enough, was Brunhild's mother and, therefore, Ingund's grandmother as well as step-mother-in-law. Google Scholar

⁶² Loc. cit. (244–5 K-L; 2.209–10 Dalton): ‘Now Leuvigild gave Hermangild and Ingund a city in which to dwell in royal state…. Ingund began to enjoin her lord … [to] forsake … his heresy…. For a long time he refused, but was moved at length by her exhortations…. When Leuvigild heard this, he began to seek occasion to bring his son to ruin. But Hermangild learned his intent, and went over to the emperor's side, entering into friendly relations with his prefect, whose army was attacking Spain.’ Cf. John of Biclar, Chronica under 579 (215 Mommsen); Isidore of Seville, Historia Gothorum (ed. Mommsen, Th. : MGH Auct. ant. 11 [n. 33 supra] 287); Leclercq, H., L'Espagne chrétienne (2nd ed. Paris 1906) 254-9; Goubert, ‘Byzance et l'Espagne,’ op. cit. (n. 1 supra) 19–29. Nearly every text relevant to Hermenegild's revolt is conveniently grouped in the recent Fuentes de la época Visigoda y Bizantinas, ed. Grosse, R. (Fontes Hispaniae antiquae, edd. Schulten, A. and Pericot, L. : fasc. 9 [Barcelona 1947] 161–94), which also includes monumental evidence. However, Grosse's commentary on passages of Gregory of Tours often requires serious revision. Outstanding for this entire period is Stroheker, K. F.'s ‘Leowigild: Aus einer Wendezeit westgothischer Geschichte,’ Welt als Geschichte 5 (1939) 446–85, specifically here 458–60.Google Scholar

⁶³ The testimony of Gregory of Tours is conclusive: ‘The unhappy prince knew not that the divine judgment impends over him who nurses such intent against his own sire, even if he be a heretic.’ HF 6.43 (315 K-L; 2.275 Dalton). Google Scholar

⁶⁴ Ziegler, Aloysius, Church and State in Visigothic Spain (Washington 1930) 30, contends that ‘… the Spanish Church did not espouse the cause of the Catholic prince against his Arian father.’ His position is unimpeachable as regards the attitude of the Spanish sources, all of which were written after the end of the rebellion, and also after the adoption of Catholicism as the Visigothic state religion. It would clearly not be in the state's interest to have a usurper glorified. At the same time, Stroheker is doubtless right in stating (op. cit. [n. 62 supra] 468): ‘Mag auch die katholische Kirche Spaniens keineswegs in ihrer Gesamtheit in dem rebellischen Königsohne den Vorkämpfer ihres Glaubens gesehen haben, so mussten doch die Katholiken die natürlichen Verbündeten Hermenegilds sein.’Google Scholar

⁶⁵ See the quotation in n. 62 supra. Google Scholar

⁶⁶ HF 4.8 (140 K-L; 2. 121 Dalton); Isidore of Seville, Historia de regibus Gothorum 46 (PL 83. 1070); Lambert, A., ‘Athanagild. 1,’ DHGE 4 (Paris 1930) 1297–1301; Stroheker, ‘Leowigild’ 448 and n. 2; Stein, Hist. du Bas-Empire II 560–4; Goubert, ‘Byzance et l'Espagne’ (n. 1 supra) 6–11. Athanagild, a Visigothic noble, rebelled against King Agila and declared himself king before December 551 (Karl Zeumer, ‘Die Chronologie der Westgotenkönige des Reiches von Toledo,’ Neues Archiv 27 [1901] 419). Unable to overthrow Agila, Athanagild appealed to Justinian for help, and the Emperor promptly sent him a Byzantine army under a good general (552). Agila's entourage was properly frightened by the arrival of Justinian's troops, and assassinated Agila so as to cut short the war and prevent the Byzantines from further advancing into Spain. Athanagild was proclaimed king and immediately turned upon his Byzantine allies; he spent the rest of his reign trying to drive them out, but the imperial army painfully managed to hold a coastal strip in southern Spain. As opposed to Stroheker and Stein, Goubert maintains the traditional date 554.Google Scholar

⁶⁷ Stroheker, ‘Leowigild’ (n. 62 supra) 479. Google Scholar

⁶⁸ HF 5.41 (248 K-L; 2.212–3 Dalton). Google Scholar

⁶⁹ HF 5.43 (249 K-L; 2.214 Dalton). Chilperic's daughter may already have been betrothed to Leuvigild's second son: supra n. 51. What the Visigothic envoy of 580 said is not known, but he probably resumed the marriage negotiations, since in 581 Chilperic sent envoys to Spain to examine the dowry: HF 6.18 (287 K-L; 2.251 Dalton; cf. n. 163 infra). Grosse, Font. Hisp. ant. (n. 62 supra) 9.183, alleges that at the outset of Hermenegild's revolt ‘… los reyes francos Gunthchramno … y Chilperico …, prepararon la guerra para amparar a Ingunda y conquistar los territorios gálicos de los visigodos.’ No evidence supports this statement. Rather it appears to me that Leuvigild sent envoys to Chilperic specifically because he knew that they would be well received. Moreover, Chilperic and Gunthram were at arm's length since 576 (n. 46 supra); and although Gunthram's aggressive designs may have been feared by Leuvigild, no testimony shows that Gunthram was actually ‘preparing war’ as early as 580, or indeed until 585. Google Scholar

⁷⁰ As supra, n. 68. Although it would seem from the order of Gregory's text that the events of ch. 41 occurred before those of ch. 43, the inversion is easily explained by the fact that Gregory placed the second chapter out of order to allow for a more extended account of his theological debate with the Visigothic ambassador. I cannot understand how Grosse (loc. cit. 156) came to date HF 5.41 ‘Hacia 574 p.C.’ Aside from the fact that the year 580 is clearly begun at HF 5.33, Chilperic in 574 did not even control Poitiers, the place where Mir's ambassadors were arrested. Google Scholar

⁷¹ Stroheker, ‘Leowigild,’ 479: ‘Die Gesandtschaft nach Konstantinopel, die zwischen 579 und 582 Leander übernahm …’ Unconvinced, Ziegler (n. 64 supra) 30, states that Leander's journey to Constantinople need not have been in Hermenegild's behalf; he is justified in the light of a strict reading of the sources. Still, I agree with Goubert's contention (‘Byzance et l'Espagne’ [n. 1 supra] 26–9) that Leander was in all probability Hermenegild's ambassador; cf. n. 64 supra. On the other hand, Goubert claims that Leander was sent to Byzantium to seek imperial intervention: ‘… c'est Herménégild qui, par l'ambassade de Léandre, provoquera son [the emperor's] intervention en Espagne….’ (Byzance et les Francs [n. 1 supra] 63). But Gregory of Tours points out that Hermenegild began his revolt in 579 when he ‘… went over to the emperor's side, entering into friendly relations with his prefect, whose army was attacking Spain’ (loc. cit. n. 62 supra). Google Scholar

⁷² Stein, , Studien 109–13; Stroheker, ‘Leowigild’ 479–80; Goubert, Byzance et l'Orient 59, 78–84. In 579 the conflict with the Avars took a particularly disastrous turn, while negotiations for peace with Persia were interrupted by the death of Chosroes I.Google Scholar

⁷³ See supra n. 66. Google Scholar

⁷⁴ HF 5.40 (247 K-L; 2.212 Dalton). Dalton's n. 5 on the same page states that this envoy died 19 August 580. Google Scholar

⁷⁵ John of Biclar, Chronica under 580 (216 Mommsen): ‘Per hanc ergo seductionem plurimi nostrorum cupiditate potius quam impulsione in Arrianum dogma declinant.’ Cf. Goubert, ‘Byzance et l'Espagne’ (n. 1 supra) 23–4: ‘… les mesures pour faciliter l'apostasie n'eurent pas grand succès.’ But in an altogether remarkable study of Leuvigild's religious policy, Stroheker concludes that conversions to Arianism were very numerous: ‘Leowigild’ 472–6, specifically here 476; Grosse, Font. Hisp ant. 9.181, comes to the same conclusion. Google Scholar

⁷⁶ For so energetic a monarch as Leuvigild to take five years to eliminate this rebellion seems to me only attributable to his wish to see the trouble disappear without actually having to come to battle with his son; Stroheker likewise believes that ‘den offenen Kampf mit seinem Sohne suchte Leowigild so lange als möglich zu vermeiden’ (loc. cit. 461). Google Scholar

⁷⁷ HF 6.1 (265–6 K-L; 2.233 Dalton). Wandelen was probably a figurehead, since he only reappears to have his death mentioned late in 585: HF 8.22 (389 K-L; 2.347 Dalton). Google Scholar

⁷⁸ Since Gogo was not given a successor until his death, his overthrow is symbolic, for the alliances were reversed before he ceased to be Childebert's governor; but this only reinforces the significance to Austrasian politics of the death of Chilperic's heirs. Would it be going too far to suggest a relation between the reversal and Gogo's death? He must have been quite old in 581. Google Scholar

⁷⁹ Marius of Avenches, Chronica under 581 (239 Mommsen; n. 40 supra): this is an index of the high importance attributed to this event in Burgundy, for Marius never bothers with minor occurrences. Gregory gives none of Mummolus’ reasons for fleeing Burgundy; I have above (n. 54) suggested some incidents which, in the light of Mummolus’ flight, might be significant; granting previous dissatisfaction, the new régime in Austrasia gave Mummolus his chance to make good an escape. Goubert (Byzance et les Francs 24) asserts that ‘… il faut voir l'action de Tibère dans l'abandon de Gontran par Mummolus….’; this supposition has no support from the sources. Google Scholar

⁸⁰ Byzance et les Francs 21-2. For Pelagius’ letter, see 82 and n. 37 supra. Google Scholar

⁸¹ Stein, , Studien 108: ‘… als sie 581 nach einer Abwesenheit von drei Jahren mit reichen Geschenken des Kaisers heimkehrte, schloss Chilperich mit den Austrasiern, den bisherigen Freunden Guntrams, gegen diesen ein Bündnis….’Google Scholar

⁸² HF 6.2 (266–7 K-L; 2.233f. Dalton); this chapter is quoted in extenso by Goubert, Byzance et les Francs 23. Google Scholar

⁸³ Byzance et les Francs 24; Stein also draws this unwarranted conclusion, n. 81 supra. Attention might be given to a curious passage which at least shows that the idea of making purchases at Constantinople existed in Merovingian times: ‘Aliquantulum solidos tuae instantiae locum accipiens militavi; parum … argentum habeo. Vellebam cum tuis legatis puerum dirigere ut melius Constantinopoli mihi argentum mercaret.’ Pseudo-Fredegar, Chronic. 3.11 (ed. Krusch, B., MGH Script. rer. Merov. 2 [Hannover 1888] 96); for comment on this passage, see n. 172 infra.Google Scholar

⁸⁴ HF 6.3 (267 K-L; 2.234 Dalton). Google Scholar

⁸⁵ Although Gregory would ordinarily treat the two events as unrelated, he would during his stay at Nogent have noted other effects of the ambassadors’ return, had there been any. Google Scholar

⁸⁶ As supra n. 84: ‘… there they discussed a plan for depriving King Gunthram of his kingdom and forming an alliance against him. And King Chilperic said: “So great were grown my sins that I have no sons more, nor remaineth there an heir to me, save only King Childebert, my brother Sigibert's son. Therefore let him be heir of all that my labors may win; only let me keep the whole for the term of my natural life without trouble or dispute.” They rendered thanks, signed a pact in confirmation of the terms discussed, and returned to Childebert with great gifts.’ It appears as if only the succession treaty was concluded and that in fact the alliance against Gunthram was not agreed upon until 583; see infra n. 149. Goubert (loc. cit.) prematurely asserts: ‘L'accord de Chilpéric et de Childebert II avait eu pour but la lutte commune contre Gontran.’ This leads him to run 581 into 583; see n. 89 infra. Google Scholar

⁸⁷ The possibility that this embassy was interested in Gundovald is discussed infra 98–9 and nn. 108–12. Google Scholar

⁸⁸ HF 6.4 (267–8 K-L; 2.234–5 Dalton). The events at Marseilles are not what they seem in Gregory's account, HF 6.11 (280–2 K-L; 2.245–7 Dalton), but should be considered in relation with the testimony of HF 9.11 (426 K-L; 2.381 Dalton) and HF 6.7 and 4.43 (276–7, 177–8 K-L; 2.242, 152–3 Dalton). Their importance for this study lies only in the fact that Bishop Theodore of Marseilles was firmly linked with the Austrasian notables and became odious to Gunthram of Burgundy. Google Scholar

⁸⁹ HF 6.12 (282–3 K-L; 2.247–8 Dalton). Under the heading ‘La révolution austrasienne (581),’ Goubert (Byzance et les Francs 24–5) runs these incidents into those of 583, which does not prevent him from later (34 n. 1 and 36) basing himself on the events of 583 to allege that Gundovald landed in that year. Google Scholar

⁹⁰ HF 6.18 (287 K-L; 2.251 Dalton). Google Scholar

⁹¹ HF 5.3, 13 (196, 207 K-L; 2.169, 179 Dalton). Cf. n. 86 supra: ‘“… only let me keep the whole for the term of my natural life….”’ When the ‘minor populus’ did overthrow the notables in 583, the cry arose: ‘“Away out of the king's presence with those who sell his kingdom, subject his cities to another master, and deliver our prince's people to foreign rule”’ (HF 6.31 [2.262 Dalton; 301 K-L]). Google Scholar

⁹² It should be emphasized that the ‘Pro-Neustrians’ felt no loyalty to Chilperic. The party included Duke Guntram Boso (who will presently be introduced), a man whom Chilperic had personal reason for detesting and had tried in the past to have executed: HF 4.50, 5.4, 14 (187, 198–9, 211 K-L; 2.159–60, 171, 182 Dalton). Google Scholar

⁹³ HF 5.14 (210 K-L; 2.181 Dalton); ‘At this time [576] Guntram sent one of his servants to a certain woman having a spirit of prophecy, known to him from the days of King Charibert.’ This seems to be more than a chronological reference, since after all Charibert acceded at the same time as Chlothar I's three other sons. It is important to differentiate the two Guntrams, the king of Burgundy and the Austrasian duke. Google Scholar

⁹⁴ HF 5.20 (228 K-L; 2.196 Dalton): ‘… irrespective of their mother's birth, all male children begotten by kings are called king's sons.’ But it first has to be proved that a king did beget the boy. Google Scholar

⁹⁵ For the plot's progress until Gundovald's arrival at Avignon: HF 6.24 (291 K-L; 2.255 Dalton). The strategy of the uprising against King Gunthram can be deduced from the actual revolt which took place in 584–85: HF 7 passim. It would be unreasonable to suppose that Egidius and the notables shared with Chilperic their plans against Gunthram. At best, they may have hoped that Chilperic would ignore an attempt by a pretender on his enemy, the King of Burgundy. Although the plotters had no reason to suppose that Chilperic would look kindly upon Gundovald, they probably calculated that he, lacking an heir of his own blood, would overlook the matter, provided he were not himself threatened. It is possible, however, that the Austrasians even intended to use Gundovald against Chilperic, as it appears that a high Neustrian official was in league with Mummolus ever since 582: HF 7.10 (332 K-L; 2.292 Dalton): ‘He [Duke Desiderius of Toulouse] made all haste to Mummolus with whom he had entered into alliance two years before [i.e. 582].’ Google Scholar

⁹⁶ From Gogo's letter to Grasulf (77 and n. 16 supra) and by the testimony of Menander (81 and n. 34 supra), it is clear that the Byzantines had previously tried to obtain Frankish aid in Italy by means of subsidy. The attempts mentioned in Menander would only have been made about three years earlier. The letter to Aunacharius (82 and n. 37 supra) testifies to even later solicitations. Google Scholar

⁹⁷ HF 6.26 (293 K-L): the -us ending is an alternative for -os in Gregory, cf. n. 170 infra and Bonnet, Max, Le Latin de Grégoire de Tours (Paris 1890) 355–6. The circumstances in which Gunthram made this accusation are not relevant to this article; see n. 140 infra. See also HF 7.32 (n. 129 infra).Google Scholar

⁹⁸ Guntram Boso's family lands were probably in Auvergne: HF 6.26 (293 K-L; 2.256 Dalton). On his return from Constantinople, he escorted his wife and sons back from Childebert's court (ibid.). Guntram's reputation for trustworthiness was not good, HF 5.14 (211 K-L; 2.182 Dalton): ‘he was too prone to break faith, and never took an oath to any of his friends which he was not at the first moment ready to break.’ On the other hand, Duke Guntram was genuinely devoted to his family, see n. 139 infra. Google Scholar

⁹⁹ In Byzance et les Francs, Goubert does not reiterate the argument he put forth in ‘L'aventure de Gondovald’ (op. cit. [n. 1 above] 423), according to which Guntram Boso would have left Gaul in 579. His earlier opinion was based on the untrustworthy evidence of Aimoin's De gestis regum Francorum. Google Scholar

¹⁰⁰ HF 7.14 (336 K-L; 2.296 Dalton): ‘“… an adventurer whose father managed a mill; if the truth must be told, he also sat over reeds and wove wool.”’ King Gunthram, who is speaking at this point, probably reflects Chlothar I's opinions concerning Gundovald's parentage. Google Scholar

¹⁰¹ Jean Hoyoux, ‘Reges criniti: chevelures, tonsures et scalps chez les Mérovingiens,’ Revue belge de philologie et d'histoire 26 (1948) 496, shows that this long hair is not the sign of royal birth, as usually claimed (and still by Goubert, Byzance et les Francs 29). He distinguishes three methods of hair removal among the Merovingians, one of which, called ‘tunděre’ by the sources, allegedly consists of violently beating the scalp so that the victim remains not only hairless but horribly scarred for life (506–8). However, Gregory uses ‘tunděre’ to describe what Chlothar caused to be done to Gundovald, which undercuts Hoyoux’ argument, for Gundovald's hair is said in the same chapter to have grown back again. Hoyoux does not cite HF 6.24. Google Scholar

¹⁰² HF 6.24 (291–2 K-L; 2.254–5 Dalton). Google Scholar

¹⁰³ HF 7.36: (Gundovald is being mocked by a hostile army) ‘“Art thou that painter fellow who in the days of King Lothar used to daub the walls and vaults of oratories?”’ (2.313 Dalton; 357 K-L). Google Scholar

¹⁰⁴ Loc. cit. (2.314 Dalton; 257–8 K-L). Gundovald is replying to the same hostile army. For the nature of this speech, see 99 and n. 116 infra. Conjecture can add little amplification to Gundovald's words. He may have fled Gaul upon learning of Charibert's death in 567, probably fearing for his life now that his last protector had gone. The mention of Narses dates the flight no later than 568, since in that year the great general was removed from his command and left for southern Italy (Agnellus in Consularia Italica, ed. Mommsen, Th.: MGH Auct. ant. 9 [Berlin 1892] 335). It might be inferred that Gundovald's wife died giving birth to his second son, and that, as a result, he decided to leave the peninsula. His mention of ‘the emperors’ lets it be supposed that Justin was reigning when he reached Byzantium.Google Scholar

¹⁰⁵ Continuatio Havniensis Prosperi (ed. Mommsen, Th. : MGH Auct. ant. 9.337–8; Paul the Deacon, Historia Langobardorum 2.28–30 (87–9 Bethmann-Waitz; 81–5 Foulke). On Byzantium and pretenders, see Diehl, , op. cit. (n. 32 supra) 204. Albsuinda was the granddaughter of King Chlothar I.Google Scholar

¹⁰⁶ See 76 and nn. 9a and 10 supra. On the methods of Byzantine diplomacy in the sixth century, see the admirable section in Charles Diehl, Justinien et la civilisation byzantine au VI ^e siècle (Paris 1901) 370–7. Google Scholar

¹⁰⁷ Agnellus, , loc. cit. (n. 104 supra); Paul the Deacon 2.2 and 3.9 (72–3, 97 Bethmann-Waitz; 54–5, 102–3 Foulke).Google Scholar

¹⁰⁸ Stein, , Studien 108: ‘So ist die Vermutung nicht allzu kühn, dass der Kaiser zur Entsendung des Gundovald zuerst durch die Gesandtschaft angeregt wurde, die Chilperich im Jahre 578 nach Konstantinopel geschickt hat….’ He then states that the treaty between Chilperic and the Austrasians contained ‘… den ausgesprochenen Zweck, ihn zu stürzen, und de[n] wenigstens von unserer Quelle Gregor von Tours nicht ausgesprochenen, aber gleichwohl erwiesenen, Gundovald an Guntrams Stelle zu setzen, beträchtliche Teile von dessen Gebiet aber sich selbst anzueignen.’ Cf. Goubert, Byzance et les Francs 22.Google Scholar

¹⁰⁹ Supra 92-3 and nn. 80–6.Google Scholar

¹¹⁰ This belief can only be supported by the silence of the texts (n. 111 infra) and by the following considerations: Gundovald had as much of a claim against Chilperic as against any other Merovingian monarch, perhaps more claim, since Gundovald might pass as heir to Charibert, whose kingdom had principally fallen into Chilperic's control; also, Chilperic is never known to have taken a diplomatic initiative: the Visigoths and the Austrasian notables always came to him, and, as I have earlier indicated, his embassy to Byzantium looks very much like a shopping expedition. Google Scholar

¹¹¹ Stein, , Studien 115 n. 6, offers only one connection between Neustria and Gundovald, to wit, HF 7.39 (362–3 K-L; 2.318 Dalton): ‘At that time Queen Fredegund dispatched Chuppa to the territory of Toulouse to bring her daughter thence in whatever manner he could. There were many who declared that he was really sent to entice Gundovald, if he still lived, with many promises, and so bring him to the queen.’ This incident takes place in the spring of 585, several months after Chilperic's murder. In view of Fredegund's precarious position at that time, the passage is more useful to show that there was no earlier connection between Neustria and Gundovald, nor, for that matter, a later one, since Gundovald was already dead when Chuppa reached Toulouse. Goubert (Byzance et les Francs 61–2) makes too much of this episode.Google Scholar

¹¹² Desiderius, Chiperic's duke of Toulouse, was in the conspiracy since 582 (n. 95 supra). Google Scholar

¹¹³ “‘I was received right graciously by the emperors….”’ HF 7.36 (n. 104 supra). Google Scholar

¹¹⁴ Gundovald's age is difficult to determine, but if he was already daubing the walls of churches in Chlothar Fs lifetime (n. 103 supra), he must then have been nearly full grown, perhaps twenty years old, at Chlothar's death in 561. On the length of his stay at Constantinople, there is the testimony of Gregory of Tours: ‘After a long lapse of years … he landed at Marseille….’ (HF 6.24 [2.255 Dalton]). The character delineation which Goubert offers (Byzance et les Francs 62–3: ‘Plus intellectuel que guerrier, plus artiste qu’énergique … etc.’; also 29) rests on the thinnest evidence. Google Scholar

¹¹⁵ HF 7.36 (358 K-L; 2.314 Dalton). Google Scholar

¹¹⁶ Goubert, ‘L'aventure de Gondovald’ (n. 1 supra) 424 n. 1, remarks: ‘Gondovald a-t-il absence de mémoire ou commet-il un mensonge?’; Gasquet, op. cit. (n. 1 supra) 185 n. 1, is also troubled by this passage. The fact that Gregory of Tours is the real author of the speech is probably the best explanation. Google Scholar

¹¹⁷ HF 7.36 (n. 115 supra): ‘“Come; all the chief men of King Childebert's realm call for thee, and not one hath dared mutter a word against thee.’”’ Gundovald quotes Guntram Boso. Google Scholar

¹¹⁸ Since the emperor could scarcely hope to restore Byzantine rule in Gaul, this is the only possible motive which would have led Tiberius to finance Gundovald. Hartmann II 1 (n. 30 supra) 62: ‘Es scheint also in der That, dass der austrasische Hof mit dem Kaiser unter einer Decke spielte und dass die Unternehmung des Prätendenten in den Augen des Kaisers den Zweck hatte, die Kräfte Burgunds … von Austrasien abzulenken und so Childebert bei seinem italienischen Zuge Luft zu machen.’ Google Scholar

¹¹⁹ It should be noted that despite the several hundred pages which have been written since the eighteenth century on Byzantium and the Gundovald affair, there is no positive evidence that the Emperor furnished Gundovald with funds, which indeed is the crucial connecting point between Byzantine diplomacy and Gundovald's adventure. What is known is that Gundovald brought to Gaul a large treasure: HF 6.24, 26; 7.36, 38 (291–3, 358, 360–2 K-L; 2.255–6, 314–6 Dalton); and there appears to be no other source for this wealth than the imperial treasury. Google Scholar

¹²⁰ This subsidy is mentioned in HF 6.42 (314 K-L; 2.274 Dalton) and will be discussed at length below. Google Scholar

¹²¹ HF 7.36 (n. 115 supra): ‘“I gave him many gifts, and in twelve sacred places received his oath that I might enter this realm in safety.”’ Google Scholar

¹²² This conclusion seems indicated by evidence that the sons were in Spain in 589: HF 9.28 (446 K-L; 2.399 Dalton). They would presumably have been sent to refuge there when the Gundovald uprising of 584–85 began to fail. Google Scholar

¹²³ The lunar eclipse is marked at HF 6.21 (289 K-L; 2.253 Dalton); Gundovald's landing: HF 6.24 (291 K-L; 2.255 Dalton); the beginning of 583: 6.25 (292 K-L; 2.256 Dalton). This chronology of events was established by Nicolas Fréret, ‘Éclaircissements sur quelques points de l'histoire de Gondevald, fils naturel de Clotaire I^er,’ Mém. Acad. Inscr. 21 (1754) 95. Fréret, who used Petau's tables, sets the eclipse at 17 September, but more recent calculations advance the date by one day: T. von Oppolzer, Canon der Finsternisse (Denkschriften Akad. Vienna, Mathematisch-Naturwissenschaftliche Classe 52 [1887]) 352 No. 2764. Although Goubert takes note of the established chronology of Gundovald's landing (Byzance et les Francs 34 n. 1), he adds: ‘D'autres auteurs [not cited] pensent que le prétendant n'arriva pas à Marseille avant le début de 583. Cette opinion semble préférable. Ainsi son débarquement ne précéderait que de peu la rupture de l'alliance neustro-austrasienne, événement qui explique la trahison de Gontran Boson.’ This is a circular argument based on the conjecture that the break of the Neustrian-Austrasian alliance explains Guntram Boso's defection. Gregory's chronology should be maintained and Guntram Boso's motives sought elsewhere. Google Scholar

¹²⁴ Blancard, , op. cit. (n. 1 supra) 436–7. 17 September (i.e. Fréret's date) should be substituted for the two misprints which give the date for the eclipse as 12 Sept. on p. 435 and 7 Sept. on p. 437. On the sailing season in the Mediterranean, cf. Lopez, R. S. in Cambridge Economic History 2 (edd. Postan, M. and Rich, E. E., Cambridge 1952) 332.Google Scholar

¹²⁵ HF 6.30 (298–9 K-L; 2.259–60 Dalton). For the actual date, Goubert, Byzance et l'Orient (n. 2 supra) 31–2. Google Scholar

¹²⁶ Fredegund was delivered of this child shortly after 18 Sept. 582: HF 6.23 (290 K-L; 2.254 Dalton), which suggests that her pregnancy would have been public knowledge by about June-July. Google Scholar

¹²⁷ The notables realized their peril to the extent of transforming in 583 the Neustrian treaty of 581 into an offensive alliance, in order to obtain immediate gains for Austrasia; but they were nevertheless overthrown by the ‘minor populus.’ See n. 149 infra. Google Scholar

¹²⁸ See n. 95 supra. Google Scholar

¹²⁹ HF 6.26 (293 K-L; 2.256 Dalton): ‘Duke Guntram first went back to Auvergne with the aforesaid treasures [Gundovald's], and then went to King Childebert.’ That the notables still formed Childebert's entourage can be deduced from the embassy led by Egidius of Reims in HF 6.31 (n. 149 infra). Their involvement in Gundovald's adventure is best stated in HF 7.32 (353 K-L; 2.310–11 Dalton; under 585): ‘… Gundovald … had been asked to take the throne by all the chief men of Childebert's kingdom; the request had been the special object of the visit to Constantinople a few years before by Guntram Boso, who had brought him the invitation to enter Gaul.’ Google Scholar

¹³⁰ HF 6.24 (291 K-L; 2.255 Dalton): ‘… he landed at Marseilles and was received by Bishop Theodore. From the bishop he obtained horses, and joined Duke Mummolus, who was then at Avignon….’ Also HF 7.36 (358 K-L; 2.314 Dalton): ‘“… the bishop received me with the greatest kindness, for he had letters written by the chief men of my nephew's [Childebert's] kingdom. I then went to Avignon, according to the wish of the patrician Mummolus.”’ The last passage shows how well everything had been planned. Google Scholar

¹³¹ As supra n. 126. Google Scholar

¹³² HF 6.24 (291–2 K-L; 2.255 Dalton). Gregory goes on: ‘But it is said that the bishop produced a letter signed by the chief men of Childebert's kingdom, protesting at the same time that he had done nothing of himself, but had only obeyed the commands of his lords and superiors. He was kept under guard in a cell….’ Google Scholar

¹³³ Guntram Boso's accusation was the basis for the now long-discredited claim that Gundovald represented an imperial attempt to recover Gaul. Neither can it be used to support the allegation that the emperor provided Gundovald with ‘diplomas,’ as does Goubert, Byzance et les Francs 64 n. 4. Google Scholar

¹⁸⁴ See n. 88 supra. Google Scholar

¹³⁵ As supra n. 132. Google Scholar

¹³⁶ As supra n. 129. Google Scholar

¹³⁷ HF 7.36 (358 K-L; 2.314–5 Dalton): “‘But Guntram Boso, regardless of his oath and promise, stole from me my treasures and took them for himself.” ’ This at any rate is the speech Gregory attributes to Gundovald shortly before his death. Google Scholar

¹³⁸ Fauriel, C., Histoire de la Gaule méridionale sous la domination des conquérants germains (Paris 1836) II 245–6.Google Scholar

¹³⁹ Guntram Boso is perhaps the non-royal figure most often mentioned in the History of the Franks, and his solicitude for his family is evident. Cf. Dezazars de Montgailhard, ‘La conspiration de Gondovald: récit des temps mérovingiens dans la Gaule méridionale,’ Mémoires de la Société archéologique du Midi de la France 14 (1886–89) 137: ‘… l'amour paternel fut peut-être sa principale vertu….’ Google Scholar

¹⁴⁰ What happened to Guntram Boso immediately afterwards, when he was accompanying his family back to Auvergne and was arrested by King Gunthram, is of no concern to Byzantine-Frankish relations, but will find a place in an account of the Merovingian aspects of the Gundovald affair which I hope to write in the future, cf. n. 142 infra. Suffice it to say that in December 584, Guntram Boso is in an Austrasian embassy to King Gunthram which numbers, among other notables, Bishop Egidius of Reims: HF 7.14 (334f. K-L; 2. 295 Dalton). Google Scholar

¹⁴¹ HF 6.24 (292 K-L; 2.255 Dalton). Google Scholar

¹⁴² There is no satisfactory account of the purely Merovingian aspect of the Gundovald affair. Kurth, op. cit. (n. 1 supra) 286–7, considers the pretender as most important during 583, when Gundovald was in reality not even in Gaul, and he hardly does justice to the rising of 584–5 (ibid. 289–94). If the Gundovald uprising of 584 can be linked to Chilperic's murder, the whole affair would assume great importance in Merovingian history. I believe the link can be made, especially in view of the most recent opinion on a key bit of testimony: HF 10.19 (510 n. 2 K-L). Google Scholar

¹⁴³ No new connection can be established between Byzantium and Gundovald; indeed, I doubt very much that Maurice would have accepted if an offer had been made. By means of an alliance between Brunhild and Gundovald at this point (see n. 150 infra), Goubert, Byzance et les Francs 36, covers the gap of more than a year and a half between the two phases of Gundovald's adventure and consequently sees Maurice's shadow stretching to the acropolis of Convenae (op. cit. 67), the place where Gundovald met his death (n. 193 infra). Google Scholar

¹⁴⁴ HF 6.18 (287 K-L; 2.251 Dalton). Google Scholar

¹⁴⁵ John of Biclar, Chronica under 581 (216 Mommsen). Google Scholar

¹⁴⁶ HF 6.18 (288 K-L; 2.252 Dalton): ‘When Ansovald was come to Chilperic, an embassy from Spain followed, which from Chilperic went on to Childebert, and then returned home,’ Google Scholar

¹⁴⁷ John of Biclar, Chronica under 582 (loc. cit.): ‘Leovigildus rex exercitum ad expugnandum tyrannum filium colligit.’ Google Scholar

¹⁴⁸ HF 6.43 (314f. K-L; 2.275 Dalton); John of Biclar, under 583 (loc. cit.); Isidore of Seville, Historia Gothorum MS H (op. cit. [n. 62 supra] 287); Grosse, Font. Hisp. ant. (n. 62 supra) 9.186; Stroheker, ‘Leowigild’ (n. 62 supra) 480; Goubert, ‘Byzance et l'Espagne’ (n. 1 supra) 30. Google Scholar

¹⁴⁹ In the spring of 583, Egidius of Reims led an Austrasian embassy to Chilperic which concluded an offensive alliance against Gunthram of Burgundy: HF 6.31 (299 K-L; 2.260–1 Dalton). Immediate steps were taken to put the plan into operation, but before the Austrasian army could effect a junction with Chiperic's, the Austrasian rank and file, ‘minor populus,’ rioted and drove Egidius and the dukes from the camp: ibid. (300–2 K-L; 2.261–2 Dalton). For their motives, see supra n. 91. Google Scholar

¹⁵⁰ I disagree completely with Kurth (n. 1 supra) 288: ‘Childebert était délivré, et sa mère aussi. Brunehaut, en particulier, put respirer à l'aise; après avoir subi pendant huit ans tous les affronts, elle redevenait reine en un jour.’ Brunhild was by no means secure until Jan.-Feb. 585, and in the meantime the notables seized power again; they are seen completely directing policy in Dec. 584, HF 7.14 (334–6 K-L; 2.295 Dalton), and had probably forced Brunhild out of power in a quiet palace revolution in July-August 584; see 115 and nn. 188–9 infra. Goubert states that Brunhild ‘Avec l'aide de Mummolus, qui venait de passer à son service, … conduira la politique dans un sens favorable à l'Empire byzantin, et hostile à Frédégonde et à la Neustrie. L'expédition de Gondovald lui servira grandement à cet effet’ (Byzance et les Francs 25–6, also 59–60, 36). The basis for this claim can be traced to his misdating of HF 6.26, which he would have occur after HF 6.31. Google Scholar

¹⁵¹ HF 5.38, 6.43 (245, 314–5 K-L; 2.210, 275 Dalton); Stroheker, ‘Leowigild’ (n. 62 supra) 480: in Mir's attempt to relieve Seville, the Byzantine troops had simply stood by. Google Scholar

¹⁵² Cf. 109 and n. 144 supra. Google Scholar

¹⁵³ John of Biclar, Chronica under 584 (216 Mommsen): ‘Leovigildus muros Italicae antiquae civitatis restaurat, quae res maximum impedimentum Hispalensi exhibuit.’ Founded in 205 B.C. by Scipio Africanus, Italica had not only been the first Italian colony in Spain but also the birthplace of the emperor Trajan.Google Scholar

¹⁵⁴ John of Biclar, under 584 (217 Mommsen); Stroheker, loc. cit. Google Scholar

¹⁵⁵ Stroheker 460, 480. Google Scholar

¹⁵⁶ Isidore of Seville, Hist. Goth. MS H (loc. cit. n. 62 supra); Stroheker 480. Google Scholar

¹⁵⁷ HF 5.38 (245 K-L; 2.210 Dalton). Goubert, Byzance et les Francs 46, calls this ‘une paix honorable.’ After a new attempt at rebellion, Hermenegild was martyred in April 585. Google Scholar

¹⁵⁸ HF 6.43 (316 K-L). When Gregory first mentions Ingund's captivity (leaving aside HF 5.38, which has no clear chronological relation to the events), he simply says that Leuvigild ‘left the woman with the Greeks’ (HF 6.40). The seeming contradiction should probably be attributed to the development of Gregory's information about Spain, which increased in direct relation to Leuvigild's repeated embassies to Chilperic; see nn. 161, 163, 167 infra. Google Scholar

¹⁵⁹ After some hesitation, I have come to agree with Stroheker in dating this event ‘etwa im März 584’ (‘Leowigild’ 481); although he does not discuss his dating, it seems to me that the determining text he cites is HF 6.40 (310 K-L; 2.271 Dalton): this embassy to Chilperic about 2 April 584 has instructions which could only apply if the rebellion had been put down. Cf. 109 and nn. 163, 167 infra. Google Scholar

¹⁶⁰ See 105 and n. 146 supra. Google Scholar

¹⁶¹ HF 6.33 (304 K-L; 264 Dalton). From Gregory's first mention of Hermenegild's revolt (HF 5.38, i.e. under 580) it is clear that his chronology of events will prove troublesome. However, his information in the early part of 584 can be seen to increase with the coming of embassies from Spain. HF 6.33 is the chapter in which the turn of the year (25 Dec.583) is recorded; the news from Spain may therefore have reached Gregory c. Jan. 584.Google Scholar

¹⁶² HF ibid.: ‘King Gunthram himself restored to [Childebert] the part of Marseilles belonging to him.’ (For the significance of this restoration, cf. 25 and n. 88 supra). This entry bears resemblance to HF 6.1, which is a summary of later events (21–3 and nn. 77, 82 supra). But in HF 6.41 (313 K-L; 2.274 Dalton), the peace with Gunthram has developed into an offensive alliance against Chilperic. It might therefore be inferred that the restoration of Marseilles was the first step in the rapprochement of Burgundy and Austrasia. Google Scholar

¹⁶³ HF 6.34 (304 K-L; 2.265 Dalton): ‘Once more came envoys from Spain. They brought gifts, and in conference with King Chilperic formally arranged the betrothal of his daughter to the son of King Leuvigild according to the previous understanding.’ Gregory writes nothing here about the end of Hermenegild's rebellion, but the actions of the embassy seem to me to indicate that a decision had been reached in Spain. The ‘previous understanding’ was no doubt made by the Visigothic embassy to Chilperic at the outset of Hermenegild's revolt, and may well have been that Rigunth would be wife to Recared if Chilperic prevented Frankish intervention in Spain during the course of Hermenegild's rebellion. Since the civil war was ended, Leuvigild was executing his part of the agreement by arranging the formal betrothal of his son to Rigunth. (This interpretation tallies very well with the next Visigothic embassy, n. 167 infra). Therefore, though Gregory does not record the fact, I believe it can be assumed that news of the end of Hermenegild's rebellion reached Gaul through this embassy. But no word probably came about Ingund until the next embassy. It might be ventured that HF 5.38 was written on the basis of information brought by the ambassadors of HF 6.34.Google Scholar

¹⁶⁴ Ibid. This event follows directly upon the departure of the Spanish embassy. The child, born in fall 582 (103 and n. 131 supra), was slightly more than a year old.Google Scholar

¹⁶⁵ Cf. nn. 167–8 infra. Google Scholar

¹⁶⁶ Easter 584 was 2 April: Gundlach, W., ‘Die Sammlung der Epistolae Austrasicae,’ Neues Archiv 13 (1888) 373 n. 2.Google Scholar

¹⁶⁷ HF 6.40 (310 K-L; 2.271 Dalton): ‘An envoy, named Oppila, came from Spain, bringing many gifts to King Chilperic. For Leuvigild, king of Spain, feared that King Childebert might march against him to avenge the affront offered his sister.’ In line with my argument in n. 163 supra, Leuvigild would have sent this embassy to keep Chilperic from believing that, since Hermenegild's rebellion was over, his obligations to Spain had ended. Google Scholar

¹⁶⁸ HF 6.41 (313–4 K-L; 2.274 Dalton): ‘Now when King Chilperic heard that his brother Gunthram and his nephew Childebert had made peace, and intended together to take the cities which he had occupied by force, …’ he withdrew all his treasures to Cambrai and told his officers to shut themselves up in walled towns; he also issued a blustering proclamation and acted indecisively with his army. It should be added that he was hard hit by the death of his son (n. 164 supra). What his relations then were with the ‘Pro-Neustrian’ party in Austrasia is an interesting subject for speculation (cf. 112 and n. 181 infra). Google Scholar

¹⁶⁹ Ibid. : At the end of the chapter, Gregory mentions the birth of a new son to Chilperic; elaborate precautions were taken to safeguard his life. This child, the later Chlothar II, was four months old at the time of Chilperic's death in Sept.-Oct. 584: HF 7.7 (330 K-L).Google Scholar

¹⁷⁰ HF 6.42 (2.274 Dalton; 314 K-L): ‘Ab imperatore autem Mauricio ante hos annos quinquaginta milia soledorum acceperat, ut Langobardus de Italia extruderit.’ Google Scholar

¹⁷¹ John of Biclar, Chronica under 584 (217 Mommsen): ‘Mauricius imperator contra Longobardos Francos per conductelam movet, quae res utrique genti non parva intulit damna’. Google Scholar

¹⁷² HF 8.18 (2.343 Dalton; 384 K-L): ‘Childebertus vero rex, inpellentibus missis imperialibus, qui aurum, quod anno superiore datum fuerat, requirebat….’ These 50,000 solidi are also mentioned in the seventh-century chronicle of Pseudo-Fredegar (loc. cit. n. 83 supra); this reference is one of numerous bits of authentic information about Byzantine-Frankish relations in the late sixth century woven into a legendary account of the return of Childeric (c. 460 !). I must forego for the present the critical study which alone can determine whether this interesting passage contains anything of historical value; it does appear to have attracted attention since Kurth, G.'s Histoire poétique des Mérovingiens (Paris 1893) ch. 7.Google Scholar

¹⁷³ To my knowledge, the first person to notice the obvious change of plan is Franz Steinbach, who writes: ‘Trotzdem [Gunthram's return of Marseilles] kam der von Brunichild betriebene gemeinsame Feldzug gegen Chilperich nicht zustande.’ (Das Frankenreich 34, in Otto Brandt, Handbuch der deutschen Geschichte [new ed. by Just, Leo; I part 2, Constance n.d.]).Google Scholar

¹⁷⁴ See 101 f. and nn. 123–4 supra. Maurice came to power too late in 582 to send an embassy in that year. Dölger, Regesten (op. cit. n. 18 supra) 1.10 no. 77, considers that Gregory refers to a letter sent to Childebert by Maurice, along with 5000 (sic) gold pieces c. 583; he continues: ‘Der feste punkt ist der 2. april 584, an welchem spanische gesandte von Leovigild sich bei Childebert [sic] aufhielten um einen nach dem Langobardenunternehmen geplanten zug Childeberts gegen Leovigild zu vereiteln’ He therefore believes that the Spanish expedition mentioned by Gregory in HF 6.42 (for which see 115 and n. 188 infra) is a doublet for the one mentioned in the previous chapter. This scarcely seems possible; that two expeditions to Spain were planned and that both were abandoned can be satisfactorily accounted for. Google Scholar

¹⁷⁵ Imperial subsidies were accompanied by considerable negotiations. Cf. Procopius (n. 6 supra); Gogo's letter (77 and n. 16 supra) probably is only part of a large complex of negotiations. Subsidies were usually solicited, as in the case of Gundovald and also infra 117 and n. 200. Unless a subsidy had previously been arranged — and these arrangements took time — imperial embassies came bearing gifts, but nothing like fifty thousand solidi. Google Scholar

¹⁷⁶ K-L 314 n. 3: ‘Verba [line] 7. Ab imperatore - [line] 11. reddere [the end of the chapter] a Gregorio postea addita esse, Giesebrecht coniecit.’ I am unfortunately unable from this reference to trace the origin of Giesebrecht's statement. For another instance of Gregory's using information later acquired, see n. 155 supra. In the present case, Gregory would presumably have made the addition not at the time of writing HF 8.18 but rather when the demand for the subsidy's return came in fall 584. Google Scholar

¹⁷⁷ On the other hand, not too much should be made of this phrase, since there is no way of determining Gregory's reason for using those very words. Google Scholar

¹⁷⁸ HF 6.40 (310 K-L; 2.271 Dalton): ‘She was the sister of Childebert and consort of Hermangild, son of Leuvigild, who after capturing and imprisoning his son, left the woman with the Greeks.’ Cf. n. 158 supra. For the sequence of arrival in Gaul of news about Spain, see nn. 161, 163, 167 supra. Google Scholar

¹⁷⁹ HF 8.21 (2.346 Dalton; 387 K-L); Epistolae Austrasicae 27, 44 (MGH Epistolae 3.139, 149–50); cf. n. 198 infra. Google Scholar

¹⁸⁰ That Maurice not only sent this message but added to it the subsidy of fifty thousand solidi (as might be argued from the assumption that ‘anno superiore’ in HF 8.18 is correct, cf. nn. 172 and 174 supra) seems to me extremely improbable. Chronology is a primary objection: Maurice would first have had to be informed from Spain that Ingund had been secured; then send an embassy to Metz with the unsolicited subsidy, in time for the Austrasians to make à summer campaign. Moreover, there is no textual evidence for this Byzantine embassy (cf. 110 supra). Furthermore, it is known that when Maurice did have Ingund and Athanagild, he asked for the subsidy's return, while pressing for further Austrasian attacks on Italy (HF 6.42 and 8.18; cf. 114 and n. 187 infra). Google Scholar

¹⁸¹ The ‘minor populus’ had accused Egidius of Reims and the dukes of betraying Childebert's cities to another master, i.e. Chilperic (n. 91 supra).Google Scholar

¹⁸² Gunthram's suspicions of Brunhild are most particularly expressed in HF 7.33 (354 K-L; 2.312 Dalton; for the circumstances and date, see nn. 192–3 infra). Kurth, op. cit. (n. 1 supra) 290–2, attributes this suspicion to Fredegund's alleged influence on Gunthram; his argument has justly been set aside by Ewig, E., ‘Die fränkischen Teilungen und Teilreiche (511–613),’ Abh. Akad. Mainz (1952) no. 9 p. 685 with n. 1, who however claims instead that Brunhild was involved in Gundovald's second rising. Goubert, Byzance et les Francs 59–60 and passim, endorses Brunhild's participation in Gundovald's adventure and therefore holds that Gunthram's suspicions of her were well founded. Since in reality the only grounds for tying Brunhild to Gundovald are Gunthram's suspicions (but cf. n. 150 supra), this is a circular argument. The fact that Brunhild abandoned the joint war on Chilperic would certainly have enraged Gunthram; that Gundovald's rising was the principal piece of skulduggery going on at the time Gunthram expressed his suspicion of Brunhild explains why he connected her with it.Google Scholar

¹⁸³ HF 6.42 (as supra n. 170); Löhlein, Italienpolitik (n. 9 supra) 65, restricts himself to repeating Gregory. Google Scholar

¹⁸⁴ See nn. 91, 181 supra, for the notables’ reasons for wishing to perform a patriotic act. Google Scholar

¹⁸⁵ HF 6.42 (as supra n. 170): ‘The emperor now hearing of the peace concluded with that people demanded the return of the money.’ For the time when Gregory would have written this, see n. 176 supra. Dölger, Regesten (n. 18 supra) 1.11 no. 84, says that an embassy came with this demand ‘585, ca. anfang,’ and adds (ibid. no. 85) that it should not be confused with the other embassy which came at that time (see n. 194 infra). It seems clear, however, that Maurice's demand came in fall 584; moreover, Gregory's phrasing is too figurative to justify the hypothesis of an actual Byzantine embassy at this time. Google Scholar

¹⁸⁶ Cf. HF 8.18 (as supra n. 172); and n. 196 infra. Google Scholar

¹⁸⁷ He was probably aware that the return of this sum was hardly likely; but he had nothing to lose which had not been already lost by Tiberius II. Cf. n. 180 supra. Google Scholar

¹⁸⁸ HF 6.42 (as supra n. 170): ‘… the king … commanded the assembling of an army which was to march against Spain; nevertheless, he remained inactive.’ As usual, an author for Childebert's decisions has to be supplied; since it was virtually impossible for an Austrasian army to march to Spain, given the current political situation, this order can be understood as an act of desperation on Brunhild's part. Google Scholar

¹⁸⁹ Ibid. : ‘sed hic fidus a solatiis nec responsum quidem pro hac re voluit reddere.’Google Scholar

¹⁹⁰ Chilperic's murder: HF 6.46 (319 K-L; 2.278–9 Dalton); Childebert at Meaux: HF 7.4 (328 K-L; 2.289 Dalton). Google Scholar

¹⁹¹ A demonstration of this statement awaits a complete account of the Merovingian aspects of the Gundovald affair. Cf. n. 142 supra. Google Scholar

¹⁹² HF 7.33 (353–4 K-L; 2.311–12 Dalton). Although Egidius and the notables had in December brought Austrasia to the point of war with Gunthram, HF 7.14 (334–6 K-L; 2.295–6 Dalton), the King broke through the web of confusion they had spun, summoned Childebert, declared him of age, and reimposed the succession treaty. For the date, see below. Google Scholar

¹⁹³ Gundovald made a last stand in the Pyrenees at Convenae (Lugdunum Convenarum, now St. Bertrand de Comminges), which he entered at the beginning of Lent, 11 February 585: HF 7.34 (354 and n. 4 K-L). Late January for Gunthram's meeting with Childebert is an approximation based on the order of Gregory's text. The end of the siege of Convenae and Gundovald's death came little more than two weeks after 11 Feb.: HF 7.37–8 (359–62 K-L; 2.315–7 Dalton). Convenae was not a ‘coin perdu des Pyrénées’ (Goubert, ‘L'aventure de Gondovald’ [n. 1 supra] 429); rather it had been an important trade center under the Empire (Camb. Econ. Hist. 2 [n. 126 supra] 41) and remained a stage on a main route to Spain (E. Salin, La civilisation mérovingienne I [Paris 1951] 128). Goubert, loc. cit. 430–1, bases on two coins found at Convenae (nos. 2428–9 in Prou, M., Les monnaies mérovingiennes [Catalogue des monnaies françaises de la Bibliothèque nationale, Paris 1892] 499) his argument for reaffirming the relevance of numismatic evidence to the Gundovald affair; he does not, however, offer sufficient grounds for denying Prou's affirmation (loc. cit. 499) that these coins were minted after 585, nor indeed for changing Prou's conclusions (op. cit. xxvvii) that no political implications can be drawn from the abundance of Gallic coins bearing Maurice's effigy.Google Scholar

¹⁹⁴ HF 8.18 (as supra n. 172): ‘King Childebert, pressed by imperial envoys to surrender the gold given him the preceding year, sent an army into Italy. It was also rumored that his sister Ingund had been removed to Constantinople.’ Google Scholar

¹⁹⁵ Ibid. : ‘But the commanders quarrelled among themselves, and returned without winning any advantage.’Google Scholar

¹⁹⁶ HF 8.21 (2.346 Dalton; 387 K-L). Google Scholar

¹⁹⁷ hf 8.28 (390 K-L; 2.348 Dalton): ‘Ingund, as I have written above, had been left by her consort with the imperial army. While she was being conducted to the emperor with her little son, she died in Africa and was there buried.’ For the assumption that she died at Carthage, cf. HF 10.2 (482–3 K-L; 2.428–30 Dalton).Google Scholar

¹⁹⁸ Epistolae Austrasicae 27 (139 Gundlach; n. 11 supra): (Brunhild to Athanagild) ‘… nepus carissime … cuius aspectum frequenter desidero … cum directis epistulis amabilibus illis oculis repraesentor, in quo mihi, quam peccata subduxerunt, dulcis filia revocatur; nec perdo natam ex integro, si, praestante Domino, mihi proles edita conservatur.’Google Scholar

¹⁹⁹ Reverdy, , op. cit. (n. 1 supra) 69–74, contends that Brunhild originally tried to interest Maurice in helping her establish little Athanagild as a pretender to the Spanish throne, and that this is the transaction alluded to in certain of the Epistolae Austrasicae; according to him, then, the end of attempts to repatriate Athanagild was prompted by the Austrasian entente with Leuvigild's successor Recared. Goubert, ‘Byzance et l'Espagne’ (n. 1 supra) 40, joins Reverdy in this opinion, which he believes to be supported by the fact that the letters to Athanagild (nos. 27 and 28) refer to him in the superscription as ‘Athanagyldo Regi.’ But how else is a king's son to be called in the absence of our meaning for the word ‘prince’? Reverdy, op. cit. 69: ‘La preuve que Brunehaut eût voulu se servir de son petitfils pour menacer les rois d'Espagne, c'est qu'après l'entente entre Reccared les souverains d'Austrasie n'insisteront plus pour qu'on leur rende l'enfant.’ The evidence is not strong enough to support such an affirmation, for the letters borne by Babo and Grippo are far more explicit on the purposes of the mission than the group of letters borne by the later embassy; in the latter, much appears to be left to negotiations carried on viva voce. With Löhlein, Italienpolitik (n. 9 supra) 66 n. 225, I agree that the letters to Athanagild show a continuing interest in obtaining his release.Google Scholar

²⁰⁰ Theophylactus Simocatta, Historiae 6.3.6–8 (ed. de Boor, C. [Leipzig 1887] 245). I owe this translation to the kindness of Mr. Mark Roskill, Henry Fellow at Harvard. It would be out of place to discuss here the problems of dating and interpretation presented by this chapter; I hope at a later time to show that the terminus a quo of this Frankish embassy can be placed in 592. Perhaps because Goubert once identified Bosus with Duke Guntram Boso (‘L'aventure de Gondovald’ [n. 1 supra] 417), he now considers the chapter confused and the names Bosus and Bettus ‘fort sujet à caution’ (Byzance et les Francs 90). Actually, Bosus is probably the Burgundian Duke Boso mentioned in HF 7.38 and 9.31; Goubert, unaware of Boso's existence, still confuses him with Guntram Boso (op. cit. 51). If it can be shown that the embassy might have occurred c. 592, the passage would constitute interesting documentation of the bad state of relations between Childebert and Maurice in the 590's.Google Scholar