¹ These coins are customarily described as bronze, but such analyses as are available (Hammer, J., Der Feingehalt der griechischen und römischen Münzen, Berlin, 1906, p. 141Google Scholar) show that they are of almost pure copper, and I prefer to use this term.

² For the history of Salerno and the dates of its rulers in the period that concerns me, I have in the main followed the excellent articles of Schipa, Michelangelo, ‘Storia del principato Longobardo di Salerno,’ Archivio storico per le province napoletane, xii, 1887, pp. 79–137, 209–264, 513–588, 740–777Google Scholar. For the period after 1077, the most useful work is the first volume of Chalandon, Ferdinand, Histoire de la domination normande en Italie et en Sicile (Paris, 1907)Google Scholar. Gay, J., L'Italie méridionale et l'empire byzantin depuis l'avènement de Basile I^er jusqu' à la prise de Bari par les Normands, 867–1071 (Bibl. des Ecoles françaises d'Athènes et de Rome. Fasc. 90. Paris, 1904Google Scholar), is also useful. The sources for the coinage are dealt with below, pp. 38–9.

³ Many of the drawings of coins with such readings as given by Spinelli, Sambon and others are not trustworthy, being the product of wishful thinking, but such coins do exist. Two photographic illustrations of coins with clear Latin legends are given in the Corpus Nummorum Italicorum, xviii, pl. XIX. 10, 11.

⁴ Texts of 1059–1061 cited in Sambon, A., Recueil des monnaies médiévales du sud d'Italie avant la domination des Normands (Paris, 1919), p. 61Google Scholar. See also below, p. 55.

⁵ This is not true of the last, however; see below, p. 49–50.

^5bis The grammar is odd, for Salerno is Salernum in Latin. The reading on the main series of the Gisulfus coins is SALERNV, but on the other series and on later coins on which the name appears in full it is SALERNO. Apparently the word is intended to be in the vernacular, and has changed its gender. It may also have been influenced by the well-known phrase aurea Roma.

⁶ Collezione Sambon (Sambon, G., Milan, 5 April 1897)Google Scholar. Two later sale catalogues of particular importance for the series in question are those of the first part of the Collezione Colonna (Canessa, C. and Canessa, E., Naples, 3 May 1909)Google Scholar, for which Arturo Sambon was also responsible, and that of the Collezioni Sambon-Giliberti (Canessa, , Naples, 10 December 1921Google Scholar), in which Arturo Sambon's own collection was finally dispersed. The text of the last catalogue is often unsatisfactory, and we can place full reliance only on the plates. See the strictures of Memmo Cagiati in his Miscellanea Numismatica, ii, 1921, pp. 163–164Google Scholar; the reply of Dell'Erba, L. in Bollettino del Circolo Numismatico Napoletano, 1921, fasc. iii, pp. 40–42Google Scholar; and Cagiati's account of the sale in Misc. Num., iii, 1922, pp. 26–27Google Scholar (‘Alcuni numeri d'ordine si verificarono non corrispondenti alle monete venute fuori, alquante classifiche e descrizioni si trovarono inesatte, non poche monete non rispondenti affatto ai dati denunziati nel catalogo, in quanto a sigle, ad emissioni, a grado di conservazione ed a numero di esemplari.’).

⁷ This volume, though consisting of little more than illustrations of the coins, without explanatory text, is essential, since it is so largely referred to by the CNI. The introduction, which reprints and expands an article earlier contributed to the Rivista Italiana di Numismatica, xxx, 1917, pp. 375–379Google Scholar, is a comprehensive bibliography of the subject to date, and absolves me from mentioning here the works of Fusco, Spinelli, Foresio, Engel, etc. Foresio's plates are sometimes useful, though the eccentricity of some of his attributions—he ascribes the coin with the legend Mense Augustu to the Lombard king Rothari (636–652)—has led numis matists to neglect his book unduly (Padre Foresio Gaetano, Le monete delle zecche di Salerno. Parte I, Salerno, 1891)Google Scholar.

⁸ These articles will be referred to in their appropriate places below.

⁹ For these events, see Schipa, art. cit., pp. 244—251.

¹⁰ For illustrations of the coins mentioned in this paragraph, see below, Fig. 1.

¹¹ Sambon, Recueil, pp. 45–58.

¹² See below, p. 57.

¹³ An idea of the correct size can be obtained from fig. 3 and fig. 4 (e) on pp. 48–51, below.

¹⁴ My thanks are due to Mr. M. C. Holderness, who made the drawings for me under my direction.

¹⁵ Cagiati's illustration is partly based on Foresio, pl. IV, 118, which shows such a star. This particular specimen, which is now in the author's collection, is overstruck on no. 26, and the star in question is to the right of the bust of Gisulf on the earlier coin. The word Victoria is never present. The letters ICT are shown on the specimen illustrated by Sambon, Rec., no. 126, and Cagiati has reconstructed the word from this, but the letters belong to a different striking.

¹⁶ Miscellanea Numismatica, ii, 1921, pp. 111–112Google Scholar.

¹⁷ The Colonna coin was overstruck on Cagiati, no. 28, which in its turn was overstruck on no. 26. Traces of the designs of both these can be seen in the illustration.

¹⁸ ‘Di alcune monete poco conosciute,’ in Memmo Cagíati's Supplemento all' opera “Le monete del reame delle Due Sicilie,” ii, nos. 8–9 (Aug.-Sept. 1912), p. 21Google Scholar.

¹⁹ Also the coin illustrated in CNI, pl. XVIII.5, a piece of crude workmanship with type and legend reversed from left to right. It is evidently a contemporary counterfeit. There is a similar specimen in the author's collection.

²⁰ Cagiati, nos. 72 and 73 should also in my view be attributed to Roger Borsa. They are usually ascribed to Roger II and the years 1127–1130 on the ground that they give Roger the title of COmes as well as DVX. The title of Comes would exclude Roger Borsa, and it was only in 1127–1130, before he assumed the title of king (1130), that Roger II was both count of Sicily and duke of Apulia. I am sceptical of the accuracy of the reading CO.

²¹ The designs of the Manso coinage are in great confusion, since the different strikings have not been fully disentangled from one another and Cagiati, Sambon, etc., are not fully in agreement as to which obverses pair with which reverses. A comparison of Cagiati no. 43 with Sambon, Recueil, no. 129 d, for example, shows that the cross in the field has no business there at all, since it is the X from the legend IC XC of an earlier striking. Since I am not here concerned to disentangle the Manso coinage, and will only have to discuss it in relation to the other Salernitan coins, I have simply included Cagiati no. 42 as a token of its existence. See further below, pp. 58–59.

²² The specimen in CNI, xviii, 317, no. 4 ( = Cat. Sambon, no. 467) is said to read GISVLFVS in the margin, but this is not visible in the illustration, nor was it so read by Sambon.

²³ Giliberti, L., ‘Un follaro inedito di Gisulfo I per Salerno,’ Boll. Circ. Numismat. Napoletano, xv, 1934, no. 2, pp. 23–26Google Scholar. He attributes it to the period of the restoration of Gisulf I (974–977).

^23bis There is an error in the drawing given here which was discovered too late for it to be corrected. Gisulf is shown wearing a crown with pendants when on the coin he only wears a bonnet.

²⁴ C. Prota, ‘Un inedito follaro religioso per Capua di Pandolfo Capo di Ferro e Ottone I imperatore (964),’ Ibid., xv, 1934, no. 2, pp. 16–19. Prota's drawing of it is reproduced in the CNI, pl. XII.24.

²⁵ Op. cit., pl. I, 120.

²⁶ The reading given in the CNI has inserted the DEO of no. 33 into the legend of this one, thus making confusion worse confounded. The Θ, an O with a horizontal stroke through it, is doubtful, since it has got confused with the S of La(u)s on the previous striking, and in any case such a letter—it is the regular abbreviation for obiit in Beneventan manuscripts of the time—would be nonsense in this context.

²⁷ These references are to Sambon's Recueil.

²⁸ The diagram suggests derivation from no. 24 rather than from no. 20, but this is because it does not represent the frequency with which the overstrikes occur. Overstrikes 19/20 and 35/20 are extremely common, while 20/24 is quite exceptional. It is no. 20 that forms the starting point of the second group.

²⁹ Recueil, p. 55, no. 129 b. The Byzantine coin on which it is overstruck is in Wroth, W., Cat. of the Imperial Byzantine coins in the British Museum (1908), ii, 455–457Google Scholar, nos. 14–29 (pl. LII, 9–12). Sambon ascribes it wrongly to Romanus II instead of Romanus I.

³⁰ Sambon, Recueil, p. 55, no. 129a; p. 53, no. 124a; also pp. 50–51. With regard to no. 124a, he says that the Byzantine striking is the later of the two, but on p. 51 he describes the order as being the other way about, and this is borne out by the illustration. A photographic reproduction of such an overstrike is given in the catalogue of Part IV of the Ruchat Sale (Santamaria, Rome, 11 June 1923), no. 1029.

³¹ Sabatier, J., Description générale des monnaies byzantines (Paris, 1862), ii, 235–236Google Scholar.

³² Wroth, , op. cit., ii, 554Google Scholar. He followed Schlumberger in ascribing them to the Crusaders, an opinion which their discovery in quantity in excavations in Greece and Asia Minor has now definitely disproved.

³³ Bellinger, A. R., The Anonymous Byzantine bronze coinage (American Numismatic Society, Notes and Monographs, No. 35, New York, 1928), esp. pp. 21–23Google Scholar.

³⁴ The Athenian Agora. Results of Excavation conducted by the American School of Classical Studies at Athens. Vol. ii. The Coins (Princeton, 1954), pp. 111, 114Google Scholar, Class I.

³⁵ Apart from what Sambon wrote about them, Dell'Erba devoted the best part (pp. 3–11) of an article ‘Sui follari longobardi anonimi alla leggenda “Victoria” battuti in Salerno’ (Boll. Circ. Num. Napoletano, 1925, pp. 3–16) to arguing that they dated from 974, and celebrated the restoration of Gisulf I to the throne. Such a date is quite impossible.

³⁶ Codex diplomaticus Cavensis, ed. Morcaldi, M., Schiani, M. and de Stephano, S., i (Naples, 1873), no. 64 (charter of Guaifar I of 868)Google Scholar; property at Amabilis is mentioned in earlier documents (nos. 45, 51, 52, 58), and the locality occasionally figures in later charters of the abbey.

³⁷ William of Apulia, Gesta Roberti Wiscardi, iv, 185 sq.Google Scholar (ed. Wilmans, R. in Mon. Germ. Hist., Scriptores, ix, 283)Google Scholar:

‘Advenit interea coniux, comitesque rogati.

Egregiam sobolem multo spectante Rogerum

Accessit populo, cunctisque videntibus ilium,

Haeredem statuit, praeponit et omnibus ilium,

Ipse quibus praelatus erat.’

Cf. also lib. v. 345 sq. (p. 297).

³⁸ II Regum, xii, 24–25 (in A. V. = II Samuel, xii, 24–25):

‘Et Dominus dilexit eum; misitque in manu Nathan prophetae, et vocavit nomen ejus:

Amabilis Deo, eo quod diligeret eum Dominus.’

³⁹ Laudes Regiae. A study in liturgical acclamation and mediaeval ruler worship (Berkeley, 1946), pp. 9–12, 222–230Google Scholar. The argument of this paper strengthens Professor Kantorowicz's view (p. 10) that in its original use in south Italy the formula was regarded as ‘very Norman,’ since it now appears that it did not derive from one previously employed by the Lombards.

⁴⁰ Registrum Gregorii VII, viii, 1a, 1b.

⁴¹ Wroth, Catalogue of Imperial Byzantine coins, pl. LX, 6, 7. This type is now ascribed to Michael IV (1034–1041), though I believe it to be a little, later. There is a very rare type of Alexius I (not in Wroth) with the same legend, but its rarity is such that it is scarcely likely to have influenced Guiscard's coinage.

⁴² CNI, xviii, 190, no. 1 (pl. VIII, 16).

⁴³ Dell'Erba in Boll. Circ. Num. Napoletano, 1923, p. 19, fig. 4.

⁴⁴ A number of references to specimens of the Gisulfus coins are usefully collected by Cagiati in his Miscellanea Numismatica, ii, 1921, pp. 107–110Google Scholar.

^44bis Also, as is pointed out below (p. 54), the armed figure of Gisulf on no. 31 is copied from a nomisma of the Emperor Isaac I Comnenus (1057–1059).

⁴⁵ Hävernick, Walter, Die Münzen von Köln. I. Vom Beginn der Prägung bis 1304 (Cologne, 1935), p. 69, no. 278Google Scholar.

⁴⁶ Ibid., p. 74, no. 306.

⁴⁷ Posse, Otto, Die Siegel der deutschen Kaiser und Könige von Pippin bis Ludwig den Bayern, i (Dresden, 1909), pl. 13, no. 8Google Scholar: counterseal of Conrad II on a document of 19 July 1033 showing a three towered representation of Aurea Roma with the legend Roma caput mundi regit orbis frena rotundi. Cf. also pl. 14, no. 4, and pl. 15, no. 4, both seals of Henry III (1039–1056). The really elaborate designs came only in the twelfth century. The Norman princes of Capua had a particular predilection for this type of seal (see Engel, A., Recherches sur la numismatique et la sigillographie des Normands de Sicile et d'Italie, Paris, 1882, pl. IIGoogle Scholar), and a simplified ‘building’ design is common on Capuan coins.

⁴⁸ Hävernick, op. cit., p. 80, no. 333.

⁴⁹ Ibid., p. 81, no. 338. Later coins show endles varieties of these architectural designs.

⁵⁰ It is probable that the coin designs are genuinely representational, the two side towers and the smaller towers depicting the fortifications on the water-front and the great central tower being the Rocca itself. Tradition had it that the advice to construct the fortifications had been sent to Guaifar by a friendly Muslim, before the siege of 871–872, through the intermediacy of a merchant from Amalfi trading in north Africa (‘ut omnimodis undique urbem suam rehaedificari faciat, et antemuralem illum qui est iuxta mare sine mora in altum elevet, aliam turrem in uno capite et aliam in alio, simulque et in medium non exiguas faciat’: Chronicon Salernitanum, c. 110, in Mon. Germ. Hist., Script., iii, 528).

⁵¹ Lindner, Theodor, Anno II der Heilige (Leipzig, 1869), pp. 42–43, 61–62Google Scholar.

⁵² Schipa, art. cit. (above, p. 37, no. 2), p. 561. The Empress Agnes also visited Salerno sometime in the late 1060's or early 1070's (di Montecassino, Amato, Storia de' Normanni, viii, 3Google Scholar: ed. V. de Barthomaeis, Rome, 1935, p. 343).

⁵³ Schipa, art. cit., p. 566. Gregory addresses him in the most flattering terms: ‘Tu autem ipse quantotius ad nos venire non pretermittas, qui, quantum Romana ecclesia te indigeat et in prudentia tua fiduciam habeat, non ignoras’ (Registrum, i, 2).

^53bis On some specimens it looks as if he is bare headed and has his hair elaborately dressed, but I think a flat bonnet is intended.

⁵⁴ This type of sceptre was then usual in the west. Besides those shown in manuscript illuminations, a certain number of actual specimens have survived, such as the so-called Petrusstab in the Domschatz at Cologne and the rather earlier Petrusstab at Limburg which once belonged to Archbishop Egbert of Trier (977–993).

⁵⁵ Beards were not at this time customary in the west, and to wear a beard was regarded as very “Byzantine.” When Archbishop Alfano of Salerno returned from the east in 1062 (?) and visited Robert Guiscard, people ‘se merveilla que vint o grant barbe, comme s'il fust de Constantinoble’ (Amato, iv, 39; ed. de Bartholomaeis, p. 211).

⁵⁶ Amato, iv, 37 (ed. de Bartholomaeis, p. 208): ‘Pour soi mostrer, porta lo vestement aorné de or et de pierrez preciouses, coment se ceste cose non se trovassent en Constentinoble, en la cort de lo Imperor. Et manda messages avant a lo Impereor, et demanda chose que jamaiz nul autre non demanda. Quar vouloit que lui fust appareillié lo siege devant lo Impereor, et fist prononcier son avenement coment ce fist un autre empereor.’

⁵⁷ Ibid., iv, 39 (p. 211): ‘Et retorna riche de li don de li Empereor.’ Earlier in the chapter the is given as 60 centenaria of gold, but this is obviously merely rumour.

⁵⁸ Schipa, art. cit., pp. 557–558, but it is not certain.

⁵⁹ Wroth, op. cit., pl. LX, 13.

⁶⁰ Some specimens of this are said to have on the obverse the legend Gisulfus prin (Sambon, Repertorio, no. 522 = Cat. Sambon 445, 446). This may be correct, but it is possibly due to over striking on no. 26.

⁶¹ There are a number of instances of the use of ancient coins in this way in the eleventh and twelfth centuries. The facing lion's scalp used as a coin-type by the Normans in Sicily was copied from Greek coins of Messana and Rhegium.

⁶² The best known and most extensive series is the Irish ‘gun money’ of 1689–1690, which is dated by every month from July 1689 to October 1690 inclusive.

⁶² Recueil, p. 52.

⁶³ Sui follari longobardi anonimi alla leggenda “Victoria” battuti in Salerno’ (cited above, p. 48, n. 35), pp. 11–15.

⁶⁵ See above, p. 38, n. 4.

⁶⁶ Amato, iv, 39 (pp. 211–12).

⁶⁷ There is no indication of date, and the reference to silver coin—Amato says that ‘tant de or, tant de argent et de rame fist faire monoie de manque poiz’—is purely rhetorical, for Salerno had no coinage of this metal. It may be noted that Amato's reference in another passage (iii, 52; p. 169) to a quadrante, which has occasioned his editor some perplexity, since he assumed it to be a coin of the period, is nothing but a Biblical allusion (Matt., v, 26).

⁶⁸ Two specimens in the author's collection have a fineness of under 8 carats, and weigh 0·96 g. and 0·92 g. A rather earlier coin, with an intelligible Arabic legend, has a fineness of about 21 carats and weighs 1·02 g.

⁶⁹ See below, p. 59.

⁷⁰ Stylistically the Italie coins are closely related to the others, but I refrain from discussing them in detail since, as said above (p. 44), I am not satisfied regarding the accuracy of the reading of the legend.

⁷¹ ‘Moncte salernitane col titolo “duca d'Italia” e monete dell'insurrezione pugliese,’ in Cagiati's Miscellanea Numismatica, ii, 1921, pp. 19–21Google Scholar.

⁷² ‘In mense Febr. factus est Argiro Barensis princeps et dux Italiae’ (Lupus Protospatarius, Annales, a. 1042, in Mon. Germ. Hist., SS., v. 58). As Greek governor of Apulia he was μάϒιστροϛ καὶ δοῦζ ἰταλὶαϛ (Chalandon, , op. cit., i, 110, n. 3Google Scholar).

⁷³ ‘Dux et magister Italiae comesque Norman norum totius Apuliae et Calabriae’ (Chalandon, loc. cit., from a charter).

⁷⁴ Engel, Recherches sur la numismatique et sigillographic des Normands, pl. I, 1. Cf. Chalandon, F., ‘La diplomatique des Normands de Sicile et de l'ltalie méridionale,’ Mélanges d'archéologie et d'histoire (Ecole franc, de Rome), xx, 1900, pp. 189–190Google Scholar.

⁷⁵ The formula is rendered into Latin, on a coin of William II of Sicily, as Operata in urbe Messana (Sambon, Repertorio, no. 1002).

⁷⁶ ‘Induzioni circa un follaro di Ruggiero II normanno in unione con Fulco di Basacers probabilmente battuto a Capua (1134?),’ Boll. Circ. Num. Napoletano, xiii, 1932, no. i, pp. 5–15Google Scholar. The form of the name in charters is Fulco de Basugerio I have failed to discover what locality in Normandy is represented by Basugerio. It might be the fairly common French place-name Bazoches in one of its various forms. Chalandon renders him ‘Foulques de Bassenger.’

⁷⁷ He illustrated the overstruck coin on p. 10, fig. 7, but few details of the earlier coin seem to have been visible.

⁷⁸ Above, p. 44, n. 20.

⁷⁹ Sambon, art. cit. (above, p. 56, n. 71), pp. 20–21. He specifically suggests the revolt of 1064, when what Amato terms ‘lo esperit de emulation et d'envie’ of the Norman barons, backed by Byzantine gold, raised all Apulia in arms against Guiscard. His view was endorsed by Prota in the article referred to below, n. 81, since Prota's hoard puts a date as late as 1134 quite out of court.

⁸⁰ Prota, in the article referred to in the next note, mentions that apart from a single specimen from Molfetta in Apulia all the specimens of the Fulcui coins whose provenance he knew of came from the region of Salerno. If Fulco were in attendance on Guiscard during the siege, as presumably he would be, this is not necessarily significant.

⁸¹ In his article ‘Un inedito follaro religioso per Capua di Pandolfo Capo di Ferro e Ottone I Imperatore (964)’, Boll. Circ. Num. Napoletano, xv, 1934, no. 2, pp. 16–22Google Scholar.

⁸² According to Prota, it did include a specimen or specimens of no. 36, which, although not definitely known from the evidence of overstrikes to be a derivative of no. 20 does I believe come into this category. If this is not simply a slip on Prota's part, I can only assume that some extraneous coins were added to the hoard before it reached his hands.

⁸³ Useful material on these coins is published in Carlo Prota's article, ‘Le monete di rame di Mansone Duca di Amalfi il primo di tal nome e Vicario Imperiale di Ottone II a Salerno (981–983),’ Boll. Circ. Num. Napoletano, xix, 1938, no. 1–2, pp. 14–31Google Scholar, though the author accepts the traditional ascription of the coins.

⁸⁴ Camera, Matteo, Memorie storico-diplomatiche dell'antica città e ducato di Amalfi (Salerno, 1876), i, 273, 293Google Scholar.

⁸⁵ Malaterra, Gaufredo, De rebus gestis Rogerii Calabriae et Siciliae comitis et Roberti Guiscardi ditcis fratris eius, iv, 24Google Scholar (ed. Pontieri, E. in Muratori, , RR. II. SS., new edn., v (1), Bologna, 1928, p. 102Google Scholar).

⁸⁶ Amato, iii, 44 (pp. 159–161).

⁸⁷ See my article, ‘The debasement of the bezant in the eleventh century,’ Byzantinische Zeitschrift xlvii, 1954, pp. 379–394Google Scholar, where the process is traced in some detail.