^{page 5 note a} Dardania, under the earlier Empire a part of Upper Mœsia, forms from the end of the third century a separate Province.

^{page 5 note b} C. I. L. iii. 3198 (and cf. 3199).

^{page 5 note c} C. I. L. iii. 3201.

^{page 6 note a} Floras, iv. 12.

^{page 6 note b} Pliny, H. N. xxxiii. 21. “Auram.… invenitur aliquando in snmma tellure protinus, rara felicitate: ut nuper in Dalmatia, principatu Neronis, singulis diebus etiam quinquagenas libras fundens.”

^{page 6 note c} Dr. Julius Jung, Römer und Romanen, p. 34 seqq. has collected the existing records of the Roman administration of Mines in Dacia, from which we may supplement our knowledge of the same administration in Dalmatia. The chief control was in the hands of a Procurator Aurariarum. Under him were various officers, such as tabularii, or treasurers, dispensatores, paymasters, and others. The exploitation was conducted by slaves condemned ad metalla, of whom there may have been 20,000, and by independent leguli aurariarum. Cf. Karl Gooss, Innerverhältnisse des Trajanischen Daciens, Excurs. I.—Die Goldbergwerke.

^{page 7 note a} To Petter, for instance (Dalmazien, B. i. p. 24 note), it is incomprehensible that the annual gold production of Roman Dalmatia should have been six times as great as that of modem Hungary, and that it should have rivalled in amount that of the South American goldfields. “Bedenkt man ferner dass der Bergbau zu den Römerzeiten noch auf den untersten Stufen stand, da den Römern alle Hilfsmittel der Jetztzeit wie z. B. Schiesspulver, hydraulische Maschinen, Dampfmaschinen, u. s. w. unbekannt waren.”

^{page 7 note b} “Mons fractus cadit ab sese longe, fragore qui concipi humana mente non possit. … Spectant victores ruinam naturæ. … Alius par labor, ac vel majoris impendii, flumina ad lavandam hanc ruinam jugis montium ducere obiter a centesimo plerumque lapide. Corrugos vocant, a corrivatione credo.” (Pliny, xxxiii. 21.) The word ruina, in the sense of “landslip” or “talus,” has been preserved in the form Räfein among the Germanized “Ladine” population of the ancient Rætia. The local names Runović, Runić, associated in several cases with Roman sites in Slavonic Illyria, may suggest a comparison.

^{page 7 note c} Loc. cit. “Alio modo puteorum scrobibus effoditur … vagantur venarum canales per latera puteorum; tellusque ligneis columnis suspenditur.”

^{page 7 note d} Silcarum, 1. iv. c. 7. Ad Maximum Junium:

“Quando te dulci Latio remittent

Dalmatæ montes, ubi, Dite viso,

Pallidus fossor redit, erutoque

Concolor auro?”

The idea has been borrowed by Silius Italicus (l. i. 231) and by Claudian, who applies the epithet “Pallentes “to the Bessian miners.

^{page 7 note e} “Cuniculis per magna spatia actis cavantur montes ad lucernarum lumina. Eadem mensuia vigiliarum est, multisque mensibus non cernitur dies.” Pliny, loc. cit. who proceeds to describe the risks which the miners ran from falls of rock and explosions of fire-damp. The ore was passed on from one gang to another, whole days and nights being consumed in the mere process of transmission: only the last lot of workmen saw the light.

^{page 8 note a} In the mountains of Montenegro and the adjoining Herzegovinian and South Dalmatian highlands I have observed iron lamps known as Lukijernar ( = lucernarius) of a form precisely similar to that found in Roman mines. The shape has survived in other European countries, but the remarkable thing here is that both shape and name should have been preserved amongst a Slav-speaking population. In the Ragusan dialect the name Lukijernar has also survived, but the lamps have lost the characteristic form preserved by the highlanders. I have already alluded to the significance of the survival of the “k” sound in “Lukijernar” and other similar fragments of the Dalmato-Roman provincial dialect among the present inhabitants.

^{page 8 note b} Martial, Ep. lib. x. 78.

^{page 8 note c} C. I. L. iii. 1997.

^{page 8 note d} Not. Occidentis, c. x. “Præpositus Thesaurorum Salonitarum Dalmatiæ.” Cf. C. I. L. 1992, 1993. 1994.

^{page 9 note a} Statius, Sylvarum, 1. 2; Epithalamium Stellœ et Violantillœ, v. 154 (referring to the Chamber of Venus):—

“Robora Dalmatico lucent satiata metallo.”

^{page 9 note b} Expositio totius mundi. (Geog. Lat. Min. ed. Riese, p. 119.)

“Dalmatia … ferrum habundans emittit.”

^{page 9 note c} Cassiodorus, Variarum, lib. iii. Ep. 25; Simeoni V. I. Comiti, Theodoricus Rex…. “Præterea ferrarias venas prædictse Dalmatiæ cuniculo te veritatis jubemus inquirere, ubi rigorem ferri parturit terrena mollicies, et igni decoquitur, ut in duritiem transferatur. Hinc, auxiliante Deo, defensio patriæ venit: hinc agrorum utilitas procuratur, et in usus humanæ vitæ multiplici commoditate porrigitur. Auro ipsi imperat et servire cogit locupletes constanter armatis. Convenit itaque hanc speciem diligenti indagatione rimari, per quam et nobis lucra generantur et hostibus procurantur exitia,” Cf. Ep. 26. Osuni, V. I. Comiti, Theodoricus Rex.

^{page 10 note a} Eckhel, D. N. vi. p. 445, remarks of these coins: “Sunt omnes ænei, III. formæ, etsi certum sit fodinas in his numis memoratas nobiliora etiam metalla fudisse. Ex quo argui potest istud monetæ genus in eorum stipendium qui ad opus in metallis faciundum destinati fuere percussum esse.”

^{page 10 note b} Cf. Neumann, Populorum Numismata, ii. 152. Rasche, Lex. Rei Numariœ, s. v. MET. NOR.

^{page 10 note c} Cohen, Médailles Impériales (2^me edition). Trajan, No. 183. There are other similar coins of Trajan with the legend METALLI VLPIANI, METALLI VLPIANI PANN., and METALLI PANNONICI. Another, representing on the reverse a female figure raising her robe and holding ears of corn, reads DARDANICI.

^{page 10 note d} Cohen, op. cit. Hadrian, Nos. 1516, 1517. That with a stag is engraved in the Pembroke Catalogue, p. iii. t. 91. Another, reading Dardanici, and with the reverse similar to the coin of Trajan, has on its obverse the head of Rome and the legend ROMA (Coh. No. 1514). Cohen omits to mention another type of this Emperor, of which I have a specimen, with MET. NOR. in an oak-wreath on the reverse, for METALLI NORICI. (Cf. Rasche, loc. cit. and Pembroke Catalogue, p. iii. t. 91.) Other coins of uncertain attribution read METAL, AVRELIANIS. These, like some of those reading METAL. DELM. present on the obverse a youthful head, perhaps of M. Aurelius, but without legend.

^{page 10 note e} Artemis is represented on the coins of the Illyrian Prince Ballæos and his successors struck at Pharia and Rhizon.

^{page 10 note f} Laureacum, where was a fabrica Scutaria; Carnuntum, which, though within the Pannonian border, must have depended on Noric mines for the same industry, and Sirmium the seat of a “Fabrica Scutorum Scordiscorum et armorum.”

^{page 10 note g} Not. Dign. Occidentis, c. 8. Fabrica Salonitana “Armorum.”

^{page 10 note h} C. I. L. iii. 2043. The tomb of a certain Maurentius FABRICENSIS.

^{page 11 note a} Cf. C. I. L. iii. 1981, 2026, 2087, 2107, 2108.

^{page 11 note b} FL. CONSTAN‖TI. NOBILISSL‖MO. AC BEATIS‖SIMO CAESARI ‖ COLL. FABRVM ‖ VENERIS. C. I. L. iii. 1981. (A.D. 333–7.)

^{page 11 note c} An inscription on the tomb of a Decurio Collegii Fabrum found at Salonæ (C. I. L. iii. 2107) concludes: SI QVIS AEAM ARCAM APERIRE VO(LV)ERIT INFERET DECVRIAE MEAE*XXV. Here the Decuria is evidently that of the Guild. In other instances we find a similar fine claimed by the Respublica Salonitana; at a later period by the Ecclesia Salonitana.

^{page 11 note d} Near Sinj. We are almost tempted to connect the figure of Equity on the Dalmatian Mine-Coinage with this COLONIA AEQVITATIS. Vide infra.

^{page 11 note e} Near Sebenico.

^{page note 11 f} T. FLAVIO ‖ T ․. FIL ․ TROmentina ‖ AGRICOLÆ ‖ DECurio ․ COLoniœ ․ SALonitaœ AEDili IIVIRo ․ IVRE ‖ DICundo ․ DECurio COLoniœ AEQVI‖TATIS ․ IIVIRO ․ QuinQnennali ․ ‖ DISPensatori ․ ‖ MVNICIPI ․ RIDITARVM ․ ‖ PRAEFecfo ․ ET ․ PATRONO ․ COLLegii ‖ FABRum ․ OB MERITA EIVS COLLegium ‖ FABEVM ․ EX AERE CONLATO ‖ CVRATORI REIPVBlicœ ․ SPLONIS ‖ STARVM ․ TRIBVNVS ․ LEGionis X ․ Geminœ ․ piœ Fidelis. … (C. I. L. iii. 2026.)

^{page note 11 g} C. I. L. iii. 1322: and cf. Mommsen's observations (p. 305), s. v. ALBURNUS MAJOR. The inscription itself was found at Zalatna in Transylvania, the ancient Ampelum.

^{page note 12 a} Hist. Rom. lib. lvi. C. 11: Γερμανικὸς δὲ ἐν τούτῳ ἄλλα τε χωρία Δελματικὰ εἷλε καὶ Σπλαῦνον, καίπερ τῇ τε ϕύσει ἰσχυρὸν ὃν, καὶ τοῖς τείχεσιν εὗ πεϕραγμένον, τούς τε ἀμυνομένους παμπληθεῖς ἒχον. Germanicus, starting from Siscia, as a base, took Splaunum on his way to Rætinium, the position of which is probably to be identified with the site of the newly…discovered Municipium near Bihać.

^{page note 12 b} Cf. Tomaschek, Die vorslawische Topographie der Bosna, &c. p. 12.

^{page note 12 c} It appears from two Bosnian documents of the years 1339 and 1422, that gold was exported from the country in the Middle Ages; and the Venetian geographer Negri, writing at the end of the fifteenth century, mentions the auri ramenta of the river Verbas. Gold-washings existed on the upper Lašva near Travnik in the sixteenth century. Cf. Jireček, ,Die Handelsstrassen und Bergwerke von Serbien und Bosnien während des Mittelalters. Prag, 1879, p. 42.Google Scholar

^{page note 12 d} Bosnien in Bezug auf seine Mineralschätze (Mitth. d. k. k. geogr. Ges. in Wien, 1870, p. 214seqq.)Google Scholar

^{page note 13 a} 13“Aus den Ueberresten dieses Bergbaues ersieht man deutlich dass das gediegene Gold in den Zersetzungs-produkten, nämlich aus dem Schwefelkies enstandenen Brauneisenstein (Brauneisenerz) und in den Ablagerungen enthalten war, welche aus den zerstrümmerten und durch die Flut weggeschwemmten Gebirgsmassen gebildet haben.” (Op. cit. p. 221).

^{page note 13 b} The present inhabitants have a superstition against continuing the search for gold, though the tradition of its existence is preserved by the local proverb:

“Vol se cese o zlatni stog a ljudi ne vide.”

(The ox rubs himself against the golden sheaf but folks see it not.)

^{page note 13 c} Ptolemy, Geog. lib. ii. c. 16, places the Pirustæ after the Dokleates (whose territory roughly answered to the modern Montenegro), and before the Skirtones, described by him as πρὸς τῇ Μακεδονίᾳ. From Livy's notice (lib. xlv. c. 26) we may infer that they lay inland from the Rhizonic Gulf. Velleius Paterculus (lib. ii. c. 115) speaks of their inaccessible position. Although, as their names show, Illyrian among the Illyrians, they are placed by Strabo (lib. vii. c. 5) in a Pannonian connexion along with their Dæsidiate kinsmen: and it is to be observed that Bato, the Dæsidiate chief, took the lead in the great Dalmato-Pannonian outbreak. We may therefore infer that there was some avenue of communication between the Dæsidiates and Pirustæ of South-East Dalmatia and the Pannonian lands of the Save: an avenue naturally supplied by the Drina Valley. From the fact that the Salona milestone places the Castellum of the Dæsidiates 156 miles distant we should be led to look for it on the Upper Drina. The Pirustæ, who as borderers of the Dokleates lay beyond the Dæsidiates, must therefore be sought in the mountain district beyond the Upper Drina. (See p. 38 seqq.)

^{page note 13 d} Cf. the deed of sale to “Andveia Batonis,” of half a house, “que est Alburno Majori Vico Pirustarum.” (Tabellœ Ceratœ, viii.; C. I. L. iii. p. 944.) Another deed records the purchase by Maximus, the son of Bato, of a female slave from Dasius, the son of Verso,—”Pirusta ex Kavieretio.” (Tab. Ger. vi.; C. I. L. iii. p. 936.)

^{page note 14 a} E. g., Anduenna Batonis (cf. Andveia above), Andesis Andunocnetis, Bato Annæi, &c, Bradua Beusantis, Cerdo Dasas Loni, Dasius (or Dassius) Breuci, Epicadus Plarentis qui et Mico, Liccaius Epicadi Marciniesus (cf. the Pæonian King, Lycceius), Lupus Carentis (from Cares), Masurius Messi, Planius Verzonis Sclaies, Plares (Plarentis), Plator Venetus, Veranes, Verzo (cf. the Dalmatian chief “Versus”).

^{page note 14 b} There is an extant diploma of Vespasian (C. I. L. iii. p.’ 849), NERVAE ․ LAIDI ․ F ․ DESIDIATI. The name occurs on a Salonitan inscription (2390) and may be compared with other Dalmatian forms in -erva, such as Derva, Anderva.

^{page note 14 c} Here was probably the station Ad Libros marked on the Tabula as 22 miles distant from Tilurio. There was an alternative way into the plain of Livno from Salonse viâ Æquum (near Sinj). While making the road from Sinj to Livno, Moiza found traces of the Roman way, and, cut on a rock at the top of the pass over Mount Prolog, the inscription “FLAVIVS MAXIMUS FECIT.”

^{page note 14 d} Tomaschek, Vorslawische Topographie der Bosna, &c. p. 22. The greatest caution, however, is necessary in accepting identifications of sites based on merely verbal coincidences. Prof. Tomaschek's ingenuity in this regard at times outruns his discretion. Thus, for example, he observes of Torine, a village near Travnik, “Der nahe Ort Torine ist unslawisch und enspricht einem alten Tarona.” So far from being “un-Slavonic” the word Torine is of universal use in Bosnia, and simply means a “sheep-fold”; a slender foundation on which to construct an ancient city. Again, heedless of the fact that “Bystrica” is the universal Slavonic name for clear streams (Old Sl. Bystrǔ, Serb. Bistar, cf. Miklosich, Die Slavischen Ortsnamen, s. v.), the same writer goes out of his way to seek for the Pannonian river Bustricius, mentioned by Ravennas, an Albanian-Illyrian origin from Buštre ═ bitch (Hü).

^{page note 15 a} The stations and mileage given by the Tabula after “in Monte Bulsinio” are—“VI Bistue Vetus—XXV Ad Matricem—XX Bistue Nova—XXIIII Stanecli”; after which follows “Argentaria” without any numerical indication. From evidence supplied by an inscription found at Rogatica (see p. 18), Bistue Nova appears to have been in the neighbourhood of that town, and Ad Matricem near the source of the Bosna. Hence we must seek for the position of Bistue Vetus about Konjica on the Upper Narenta, and it becomes evident that a deficiency must be supplied either in the names or mileage of the earlier stations of the Tabula

One of the Bistues, probably Bistue Vetus as being nearer to the maritime tract, seems to have been still flourishing in the sixth century. An “Andreas, Episcopus Ecclesiæ Bestoensis” is mentioned in the Act of the Provincial Council of Salona of 530 and 532 (Farlati, Illyricum Sacrum, t. ii. p. 173).

^{page note 15 b} Cf. Jireček, ,Die Handelsstrassen und Bergwerke von Serbien und Bosnien während des Mittelalters (Prag, 1879), p. 85.Google Scholar The plain of Serajevo was known as the Župa Vrchbosna, but the strong hold was on the site of the present citadel of Serajevo, not at the actual source of the Bosna as has sometimes been asserted. As early as 1436 we find a Turkish Voivode placed here to control the tributary Christian dynasts of Bosnia.

^{page note 16 a} Through Bosnia and the Herzegovina, &c. 1876, p. 237 (2nd ed. p. 237).

^{page note 16 b} Arch. Epigr. Mitth. aus Osterr. iv. 44.

^{page note 16 c} For this I am indebted to the kindness of M. Moreau, the French Consul at Serajevo, in whose hands the fragment now is. It is six inches in height. From Vitina, near Ljubuški, in Herzegovina (cf. C. I. L. iii. 6365, 6368; Hœrnes, op. cit. p. 41), the same gentleman had obtained a finger of a colossal marble statue, and a tile with the inscription LEG VIII AVG.

^{page note 16 d} The name is equivalent to “church-land.” A part of it is still used as a cemetery, and several mediæval tombs of the usual kind are to be seen, indicating the former existence of a church (crkva).

^{page note 17 a} See p. 31 seqq.

^{page note 17 b} More recently Heir Dumičić has discovered in the same neighbourhood, on the left bank of the Lepenica near Kisseljak, and not far from the confluence of the Fojnička Rjeka, the following inscription:

C. MANLIVS

C L

RONESIMVS

ANN |11‖

The cippus on which this was inscribed lay amongst hewn stones and other ancient fragments on a steep rock called Crkvice, to the north of which is a sloping terrace. (Arch. Epigr. Mittheilungen aus Oesterreich, 1883, p. 130.) A fragmentary sepulchral inscription has also been discovered by Captain Von Handel at Divjak in the Lašva valley south of Travnik.

^{page note 17 c} See Archæologia XLVIII. p. 66.

^{page note 18 a} Jireček, op. cit. p. 49. Foinica or Chvojnica is frequently mentioned in the Ragusan archives of the fifteenth century as the seat of a mining colony of the Republic which numbered amongst its members scions of the patrician houses of Bonda, Bucchia, and Gozze.

^{page note 18 b} Herr Dumičić of Kisseljak showed me specimens of ore from this neighbourhood containing as much as thirty per cent. of silver.

^{page note 18 c} Accompanied by grains of silver, cinnabar, and globules of quicksilver.

^{page note 18 d} Through Bosnia, &c. p. 210, 227, seqq.

^{page note 18 e} Rudnik is derived from the Old Slavonic Ruda ═ Metallum. Cf. Miklosich, Die Slavishen Ortsnamen aus Appellativen, s. v.

^{page note 18 f} The first describers of this inscription, Dr. Blau and M. de Ste Marie, differed as to their reading. Dr. Blau reading DEC. C. RIS completed by Mommsen (C. I. L. iii. 2766 b) Dec(urio) C(ivitatis) Ris(ini): (Itinéraires de l'Herzegovine); M. de Ste Marie reading DEC. C. BIS to be completed Dec(urio) C(ivitatis) Bis(tuae) or Bistuensis. Dr. Hoernes on first examining the stone accepted Dr. Blau's version, though with the remark that “das unten beschädigte R einem B ähnlich sieht”(Arch. Epigr. Mitth. iv. p. 45); but on a second examination of the stone in 1880 he convinced himself that the true reading was BIS. Identifying the “Mun(icipium) S.” on the site of Plevlje with the Stanecle of the Tabula, he observes that it must be the Bistue Nova, which is to be sought at Rogatica or Gorazda, and adds the obvious corollary, “Dann ist aber auch die Lage von ad Matricem bestimmt und wir müssen diese wichtige Station in das Quellbecken der Bosna verlegen” (Alterthümer der Hercegovina, ii. 139.)

^{page note 19 a} Tomaschek compares the Pannonian and Galatian “Matrica” and the “Mediomatrici” of Metz and seeks a Celtic origin. It is always possible that the Latin name was due to some adaptation of an earlier indigenous form.

^{page note 19 b} “Cf. C. I. L. iii. 996, 1152, 1284, 1285, 1364, 1555, 1599, 6313.

^{page note 19 c} C. I. L. iii. 6313. The remains of the temple and the inscription were discovered in 1865 by Dr. Janko Šafarik, and are described in Glasnik, 31, 217—236.

^{page note 19 d} C. I. L. iii. 996.

^{page note 20 a} C. I. L. iii. 2766a.

^{page note 20 b} We may include in the same category the Ragusan and Italian architects, known in several instances to have been employed by the Turkish Pashas in Bosnia, &c. to build bridges. Cf. p. 24.

^{page note 21 a} Strabo, Geogr. lib. vii.

^{page note 22 a} Vell. Paterculus, lib. ii. c. 110. “In omnibus autem Pannoniis non disciplinæ tantummodo sed linguæ quoque notitia Romance: plerisque etiam literarum usus et familiaris animorum erat exercitatio.”

^{page note 22 b} The only discrepancy that suggests itself is the non-prolongation of the cross-line of the H to the perpendicular line of the D. A parallel instance however may be found on coins of Athalaric, and it appears that in both cases the H was an approach to the so-called “Lombardic” h. We should thus read DN ThEoDoRICI.

^{page note 22 c} On the King's own seal, doubtless, THEODORICI REGIS. The signet ring of Childeric had the inscription CHILDIRICI REGIS (Chifflet, Anastasis Childerici Regis, p. 97, Antwerp, 1655). The insertion of the D.N. shows that the present gem belonged to an official and not to the king himself.

^{page note 23 a} Cassiodorus, Variarum, lib. iii. Ep. 25. See p. 9.

^{page note 23 b} Anon. Valesianus, c. 79. “Igitur rex Theodoricus illiteratus erat, et sic obruto sensu, ut in decem annos regni sui quatuor litteras subscriptionis edicti sui discere nullatenus potuisset. De qua re laminam auream jussit interrasilem fieri quatuor litteras regis habentem THEOD. ut, si scribere voluisset, posita lamina super chartam, per eam penna duceret, et subscriptio ejus tantum videretur.”

^{page note 23 c} See Archæologia XLVIII. p. 90, 91.

^{page note 24 a} See p. 18.

^{page note 24 b} A letter of the Ragusan Government to their ambassador at Constantinople, dated Sept. 19, 1568 (given by Jireček, op. cit. p. 86), refers to the construction of this bridge. “Dovete sapere che nelli mesi passati fummo ricercati dall Ill. Signor Mustaffa Bassa di Buda che li dovessemo mandare marangoni, muratori, fabri et molte cose necessarie perche sua Signoria dovea fabricare per fare elemosina un ponte in Ghorasda al quale habbiamo servito volentieri.” This Ragusan bridge was of five arches of woodwork, resting on piers of deftly-hewn stone blocks, oblong in shape but not so thick as Roman blocks. The woodwork was so constructed that the middle of the bridge was greatly elevated.

^{page note 24 c} The male peasants—less conservative in dress than their womankind—(except in Albania, an almost universal rule in the Ottoman dominions in Europe) have adopted the Oriental and Slavonic attire of the surrounding populations. In parts of North Albania the fustanella is common to both sexes.

^{page note 25 a} Cf. Jireček, op. cit. p. 73.

^{page note 25 b} Delle Cose de Turchi, p. 6. (In Vinegia, 1541.) Ramberti groups “Plevie” with Prijepolje as “secondo il paese assai grandi e buoni.”

^{page note 25 c} Cf. Blau, Monatsbericht d. k. Preuss. Akad. 1866, p. 840. He adds, “Noch jetzt wird von Plevlje über Gatzko und Trebinje ein namhafter Handel mit Ragusa getrieben.”

^{page note 26 a} Giornale del Viaggio a Constantinopoli fatto dagli Ambasciatori della Repubblica di Ragusa alla Sublime Porta l'Anno 1792. (In Engel. Geschichte des Freystaates Ragusa, Wien, 1807, p. 312, seqq.)Google Scholar

^{page note 26 b} Monatsbericht der k. Preuss. Akademie der Wissenschaften, 1866, p. 838, seqq. The inscriptions copied by Dr. Blau are given in C. I. L. iii. 6339–6357.

^{page note 26 c} Archäologisch-Epigraphische Mittheilungen aus Oesterreich, 1880.

^{page note 27 a} The inscriptions in the Podstražica Mosque are given by Dr. Blau (cf. C. I. L. iii. 6344, &c.)

^{page note 27 b} C. I. L. iii. 6344.

^{page note 27 c} C. I. L. iii. 6345.

^{page note 27 d} Op. cit. p. 7. “Im Hof derselben Moschee ist eine etwa Mannshohe Stele bis an den Fuss in die Erde vergraben. Ich konnte sie nur ein paar Fuss tief blosslegen und ueberzeugte mich, dass die Vorderseite eine römische Inschrift trägt, deren letzte Zeile die Buchstaben (M)ONVM(entum) enthält.”

^{page note 27 e} Not, as erroneously given by Blau (C. I. L. iii. 6339), I. O. M. The N is perfectly clear, and cannot be regarded as an imperfect M.

^{page note 27 f} Cf. I. O. BESSVMARVS. C. I. L. iii. 1053.

^{page note 27 g} It is given in C. I. L. iii. 6343. The punctuation, line 2, is however. E. Q. E.

^{page note 27 h} Fig. 8 is from the Podstražica mosque. Fig. 9 from the konak of Sali Beg. The inscriptions are incorrectly given by Dr. Blau (C. I. L. iii. 6346, 6349). My copies agree with Dr. Hœrnes' collations.

^{page note 28 a} The place is called Radosavac. The inscriptions are accurately described by Dr. Hoernes and need not be repeated here.

^{page note 29 a} My own impressions regarding the site will be found to agree generally with those of Herr Müller and Dr. Hoernes as given by the latter in Arch. Ep. Mittheilungen, loc. cit. I differ, however, from my fellow-explorers in considering that the ancient site extended also to the right bank of the Vežeznica. I may take this opportunity of expressing my obligations to Herr Müller for his valuable advice, although he was unfortunately absent from Plevlje at the time of my visit.

^{page note 30 a} Near here Dr. Hoernes found a fragment of an inscription reading L ….. ‖ CAMBRIANVS ‖ L. P.; apparently in situ—”Wahrscheinlich noch unverrückt an seiner ursprünglichen Stelle.”

^{page note 30 b} This reads I. 0. M. ‖ STATIVS ‖ VICTOR. BRI‖ZIDIA. V. L.ςA. The last line is not quite correctly given by Dr. Hoernes, who gives v. L. P.

^{page note 30 c} It reads I. O. M. F ‖ M. AEMIL ‖ ANTONIVS ‖ II. VIR ‖ L. P.

^{page note 31 a} One of these had been knocked away by the Turks, who recently gutted the church and burned the priest's house. I found it in the yard of a cottage at Grevo, below the hill of St. Ilija, with some other ancient fragments.

^{page note 31 b} At the village of Gomiljani the treatment of the drapery was curiously similar.

^{page note 32 a} Sabatier, Monnaies byzantines, i. 276, No. 59.

^{page note 32 b} A head of Zeus appears on some autonomous Illyrian coins of Scodra and Rhizon.

^{page note 33 a} This explanation of the practice of depositing the body in a contracted position has been suggested by my father in his Ancient Stone Implements &c., of Great Britain, p. 135.

^{page note 33 b} In C. I. L. iii. 6347, Dr. Hoernes read · ADIL · A, and considered that it contained the name FADILLA.

^{page note 35 a} This parallel has been pointed out by Dr. Otto Blau (Reisen in Bosnien, p. 64), who gives many examples.

^{page note 36 a} Blau (loc. cit.) cites among female Albanian names of this kind, Laljo, Liljo, Kondo, Drano, &c.

^{page note 36 b} Not in C. I. L. The inscription is given by Dr. Hoernes, loc. cit. p. 9. My own copy is somewhat fuller.

^{page note 36 c} C. I. L. iii. 3720, 3724, 3725, 3726, 5708, 5736, 5737, 6467.

^{page note 38 a} Ptol. Geog. lib. ii. c. 16.

^{page note 38 b} Cf. Livy, lib. xlv. c. 25. For their connexion with Lychnidus, see lib. xliii. c. 9. “(Appius Claudius) ad Lychnidum Dassaretiorum consedit.” Lychnidus was a central station of the Egnatian Way, and Pylon, a little beyond it to the East, was reckoned the boundary of Illyricum and Macedonia (Strabo, Geog. lib. vii.)

^{page note 38 c} The silver coins of Damastion throw an interesting light on ancient Illyrian and Epirote mining industry. On the reverse of some of them are represented hammers, picks, the symbol of fire, and an object which Professor Gardiner, with great probability, considers to be bellows. The exact site of Damastion remains to be identified, but Dr. Imhoof-Blumer, in his interesting account of some of the coins in the Zeitschrift für Numismatik (B. i. p. 99, seqq.), calls attention to the village name of Damesi, near Tepelen, where silver mines appear to have anciently existed. Closely allied to these coins of Damastion are those of Pelagia and others with the legend ∑APN0ATωN. The attempt of Dr. Imhoof-Blumer to identify the name Pelagia with Belagrita, an older form of the Albanian Berat, cannot be accepted, it being simply an Albanian corruption of a Slavonic Belgrad; Tomaschek's comparison with Pljage is more hopeful. With regard to the attribution of both these places, however, I shall venture some new suggestions. (See p. 89.)

^{page note 39 a} Hist. xliv. c. 31, and xlv. c. 43. Polybios, xxx. 19.

^{page note 39 b} The scene of the campaign of King Genthios’ brother against the native rebels is indicated by his subsequent capture by the Roman general at Medeon to be identified with the hill-fortress of Medun, in Montenegro. This district was then occupied by the tribe of the Dokleates, whose civic centre Doklea still survives in the modern Montenegrin village of Dukle. See Archaeologia, vol. XLVIII. p. 84.

^{page note 39 c} Prof. Stojan Novaković (Rad. xxxvii. (1876), 1–18) believes to have identified the site of this important old Serbian staple with the site of Plava, in the vale of Gusinje, where according to Hecquard are remains of a more ancient city. It is certain that Brskovo, the Brescova of the Ragusans, lay somewhere on the Upper Lim. (See Jireček, op. cit. p. 69.)

^{page note 39 d} Geogr. lib. Vii. Ἔθνη δ᾽ ἐστὶ τῶν Παννονίων…Πειροῦσται καὶ Μαζαῖοι καὶΔαισιτιᾶται.

^{page note 39 e} See p. 14.

^{page note 40 a} Strabo, for example (lib. vii.), mentions that the Romans had driven the once piratic race of the Ardiæi away from the sea to a sterile tract of the interior where in the impossibility of obtaining sustenance the whole race had almost died out. He adds as similar examples the case of the Autariatæ and Dardanii, the Gallic Boii and Scordisci, and the Thracian Boii.

^{page note 40 b} Velleius Paterculus, lib. ii. c. 115. “Quippe Daorisi et Dæsitiates Dalmatas, situ locorum ac montium, ingeniorum ferocia, mira etiam pugnandi scientia et præcipue angustiis saltuum pœne inexpugnabiles, non jam ductu, sed manibus atque armis ipsius Cæsaris, tum demum pacati sunt cum pœne funditus eversi forent.” The Daorisi, Daorsi, or Daversi had, like the Ardiæi, been a maritime people, and, as is proved by their coins representing a galley with the legend δAOP∑ωN, had shown themselves receptive of Greek culture. Their original area lay to the South of the Narenta mouth. For the Autariatæse see Strabo, loc. cit.

^{page note 40 c} Hist. lib. xlv. c. 26.

^{page 41 note a} I observe that the same etymology has occurred independently to Dr. Simo Rutar, Starine Bokokotorske (“Antiquities of the Bocche di Cattaro,” in Program c. k. realnog i velikog Gimnazija u Kotoru, 1880). “Pri brojenju ovih slobodnih obćina spominje Livij Pirustas odmah prije Risna. I dandanašnji imamo grad odmah pred Risnom, kojega ime, skoro do slova, jednako glasi kao Pirustœ, t. j. Perast ․ ․ ․ ․ od koga znamo da je prestari grad i da narod izvadja njegov izvor već iz doba rimskih carevâ.” (“In enumerating these free communities Livy mentions the Pirustæ immediately before Rhizon (Risano). At the present day too we have a town in the immediate neighbourhood of Risano the name of which corresponds almost to a letter with that of the Pirustæ, namely Perasto, … of which we know that it is a town of great antiquity, the origin of which is traced back by the people to the time of the Roman Emperors.”)

^{page 45 note a} Arch. Ep. Mitth. 1880, p. 14.

^{page 45 note b} Cf. C. I. L. iii. 2787, “PLADOMENVS ․ SERA ․ TVRI ․ F “; 2797, “VENDO TVDANIA PLADOMENI F “; 6410, “(I) ․ o ․ M APLV. DV//// MEVERTENS ․ PLαDOMENI ․ FILIV‖.” All from Municipium Riditarum. The termination domenus has a Celtic sound, e. g. Dumno-vellaunus, Dumno-Rix, Cogi-dubnus, &c.

^{page 46 note a} By the Turks called Hissardjik.

^{page 46 note b} Ramberti, Viaggio da Venetia a Constantinopoli (In Vinegia, 1541), p. 6, “Passammo il castello di Millesevatz ed il Monte Molatscidi, che è come a dire Montagna di Morlacco.”

^{page 46 note c} Presbyteri Diocleatis Regnum Slavorum (In Lucius de Regno Dalmatiœ et Croatiœ (Frankfort, 1666, p. 288): “Vulgari (sc. Bulgari) post hæc ceperunt totam provinciam Latinorum qui illo tempore Romani vocabantur modo vero Morovlachi hoc est Nigri Latini vocantur.” Opposed to these Crni Vlahi, or black “Vlachs “as they were also known, were the Bijeli Vlahi, or white “Vlachs,” but on what the distinction was founded is uncertain. At a later period Mavrovlachia appears as the equivalent of Moldavia. It is to be observed that Lucius of Traü supplies the right derivation of the word Morlach; and to him is really due the credit of having in his masterly chapter de Vlahis exploded the fallacy of their Transdanubian origin. The chief arguments adopted by Sulzer, Roesler, and other writers of recent times, will be found clearly and succinctly stated by the seventeenth-century Dalmatian antiquary.

^{page 47 note a} Both Surdulǔ and Ursulǔ occur among the Rouman personal names in the foundation charter of the church of the Archangel at Prisren, issued by the Serbian Emperor Dušan in 1348.

^{page 47 note b} Procopius de Ædificiis, lib. iv.

^{page 47 note c} These names are of peculiar value, as giving us an insight into the nomenclature of the” country districts of Illyricum in the sixth century of our era, a subject on which historians and geographers are for the most part silent. The ϕϱούϱια of Justinian were mostly small castles, or even mere blockhouses, like the later Turkish karaulas, for the protection of the country-side. The age of castle-building on peaks has begun, and the sixth-century Castellum was doubtless in many cases the local predecessor of the “Grad,” or central stronghold of the Slavonic ”Župa.” The Roman or Romance names have frequent reference to mineral and other natural sources of revenue which it was desirable to protect, as Æraria, Ferraria, Argentarias, Lapidarias; in many cases they contain an honorary tribute to Emperors and Empresses, who reigned in the fourth, fifth, and sixth centuries, e. g. Constantiana, Justiniana, Pulchra Theodora, Placidiana, &c. Names like Castelona, Braiola, Vindemiola, Lutzolo, Casyella have a decidedly Italian ring: others such as Ducepratum (?Doucepré), Lupofontana, Lucernarioburgus show us that the neo-Latin language of Illyricum had attained a Teutonic facility for forming compounds. In some instances, as “Sabini-bries” and “Prisco-pera,” Latin and Thracian elements are blended. The Thracian, Illyrian, Slavonic, and Gothic name-forms are of the highest interest, but can only be referred to here.

^{page 48 note a} Cf. Wallachian, septe = 7. Accepting Tomaschek's emendation of another name in Procopius' Catalogue, “tredecitilias “gives us already the Wallachian tredeci = 30.

^{page 48 note b} This compound reminds us of the common Slavo-Rouman local name Lupoglava =Wolf's head.

^{page 48 note c} This pass led from Illyricum into Greece.

^{page 48 note d} See Archaeologia, vol. XLVIII. p. 86.

^{page 51 note a} Archaeologia, xlviii. p. 11.

^{page 53 note a} The Laconicum, being merely a steam-bath, had no piscina, as will be seen from the representation of the chamber supposed to be a Laconicum discovered at Pisa, and given by Robortelli (in Scribonius Largus, ed. Rhodius. Patavii, 1655). This Pisan example is a domed circular chamber with niches, small square windows round the vault, and an opening at the top.

^{page 53 note b} In Athen. xi. p. 561, quoted by Marquardt, Römische Alterthümer, part v. p. 299.

^{page 53 note c} The jealous precautions of the Turks prevented me from examining the interior.

^{page 54 note a} It would be interesting to know know far the bath-buildings restored by Theodoric over the famous hot springs of Aponus, near Patavium (Cassiodorus, var. ii. Ep. 39), were the counterpart of S. Giovanni in Fonte.

^{page 54 note b} Procopius, De Ædificiis.

^{page 54 note c} De Adm. Imp. c. 32. The Bulgar Prince Blastimer, captured by the Serbs, is on his release safely re-conducted μέχϱι τῶν συνόρων ἓως τῆς Ῥάσης.

^{page 55 note a} Τὸ Ράσου ϕρούϱιον in Kinnamos (Hist. lib. ii.) taken by the Serbs from the Byzantines (Hist. lib. iii.); retaken by the Emperor Manuel. Kinnamos reckons it a Dalmatian stronghold.

^{page 55 note b} The castle of the Župans and later Kings is, as Jireček points out (Die Handelsstrassen, &c. p. 77), to be sought in the neighbourhood of the episcopal church.

^{page 55 note c} A description of the remains of Grjurgjevi Stupovi will be found in Travels in the Slavonic Provinces of Turkey-in-Europe, by G. Muir Mackenzie and A. P. Irby, 2nd ed. vol. i. p. 273.

^{page 55 note d} Jireček, op. cit. p. 77.

^{page 55 note e} Captain Sterneck of the Austrian Survey has given a very imperfect copy of a Roman sepulchral inscription from Banjska in his Geographische Verhältnisse, Communicationen, und das Reisen in Bosnien, der Herzegovina, und Nord Montenegro, Pl. IV. (Vienna, 1877).Google Scholar

^{page 56 note a} The Slavonic Provinces of Turkey-in-Europe, by G. Muir Mackenzie and A. P. Irby, vol. i. p. 262 (2nd ed.)

^{page 56 note b} Giornale del Viaggio a Constantinopoli fatto dagli Ambasciatori della Republica di Ragusa alla Sublime Porta l'Anno 1792. “In distanza di un' ora del sequente alloggio (Banjska) trovarono una fonte che scorreva in un' urna antica ben lavorata, ma molto patita, coll' izcrizione latina che per troppo fretta non ebber commodo di leggere.” (In Engel. p. 320.)

^{page 57 note a} E. g. Ramberti, Delle Cose de Turchi, p. 7 (In Vinegia, 1541): “Passamo la Montagna dell' Argento … si chiama dell' Argento perchio che continuamente vi stanno huomini in essa che cavano argento.”

^{page 57 note b} Die Handelsstrassen Serbiens, &c. p. 55. “Novo Brdo (Novaberda, Novabarda, in Lat. Urk.) Novus Mons, Novomonte der Italiener, Nyeuberge der sächsischen Bergleute, Νοβοπύργον, Νοβοπϱόδον der Byzantiner, war, 1350–1450, die grösste und berühmteste städtische Ansiedelung des ganzen Innern der Halbinsel. Von ihren Schätzen erzählte man sich im Auslande ganz fabelhafte Geschichten; der Byzantiner Kritobulos schreibt Gold und Silber werde hier förmlich aus dem Boden hervorgeackert.”

^{page 57 note c} Bertrandon La Brocquière, Counsellor and First Esquire-Carver to Philip-le-Bon, Duke of Burgundy, Travels to Palestine and return from Jerusalem overland to France during the years 1432–1433. Translated by T. Jolmes at the Hafod Press, 1807, p. 274. “The Despot of Servia possesses towards the common -confines of Bulgaria, Sclavonia, Albania, and Bosnia, a town called Nyeuberge, which had a mine producing both gold and silver at the same time. Each year it pays him more than two hundred thousand ducats, as well-informed persons assured me; without this he would be soon driven out of his dominions.”

^{page 58 note a} Since this paper was communicated to the Society of Antiquaries a copy of the following interesting inscription found at Batus, in the Kossovo Polje, has been sent by.Signor Paolo Orsi to the Arch. Epigr. Mittheilungen aus Oesterreich (1883, heft 1. p. 146), the ligatures here omitted:

I ․ O ․ M ․ V / PP

D ․ D ․ ET ․ GEN//

STATIONIS

PRO S ․ DN ․ IMP ․

[SEVERI ․ ALE

XANDRI] AVG

VALERIANVS

SPECVL ․ LEG ․ IIII

/L ․ S ․ A ․ V ․ S ․ L ․ M ․ AVG ․

SEVERO ․ ALEXAND ․ AVG.

//ET AVFID ․ MARCELLO

․ ․ ․ ․ ․ ․ ․ ․ ․

Which is there read:

J(ovi) O(ptimo) M(aximo) d(omni) d(ivinœ) et Gen(io) Stationis pro s(alute) d(omini) n(ostri) Severi Alexandri Aug(usti) Valeriannus specul(ator) Leg(ionis) IIII (F)l(aviœ) S(everianœ) A(leœandrianœ) V(otum) S(olvit) 1(ibens) m(erito) Aug(usta) (sic) Severo Alexand(ro) [II] etAufid(io) Marcello [II Co(n)S()ulibus]. The D.D. in the second line seems to connect itself with the Municipium D.D. the existence of which I have now established in this neighbourhood. Perhaps the preceding letters should be read R.P., i. e. Rei Publicæ D.D. The inscription is of 226 A.D.

^{page 58 note b} Putopis Stare Srbije, pl. i. (since published by Engelhardt, Révue Archéologique, 26 (1863), 141; Eph. Ep. ii. 500). It reads; VLP IONICE HAVE BENE VALEAS QVI ME/SALVTAS/, D. M/CLAVDIA RVFINA/ VIXIT ․ ANNIS ․ XXX/VLPIVS ․ IONICIANUS / VIXIT ANNIS XXV / VLPIVS RVFINVS / VIXIT ANNIS V ․ H ․ S ․ S / M ․ VLP ․ IONICVS CO/IVGI ET FILIS B ․ M / ET SIBI VIVVS / F ․ C.

^{page 58 note c} Handelsstrassen und Bergwerke von Serbien, &c. pp. 2, 68.

^{page 59 note a} The inscription has been published by Hilferding (Bosnia, Herzegovina i Stareja Serbia) (Eph. Ep. iv. 215) in an incorrect form.

^{page 60 note a} Reise von Belgrad nach Salonik, p. 240. C. I. L. iii. 1691. I was informed by the monks that this inscription had originally been found on Mount Veljetin above the town, where there are said to bo other remains.

^{page 60 note b} C. I. L. iii. 1695. I could no longer see 1694.

^{page 62 note a} See Jireček, Die Handelsstrassen, &c. p. 57.

^{page 62 note b} Some of the inhabitants here are recognised to be Roumans; most understand the Rouman language. Their wanderings sometimes extend beyond the Russian frontier.

^{page 62 note c} In Ravennas the name appears under the form Ulciano.

^{page 63 note a} Jordanes, De Getarum sive Gothorum Origine, c. lvi.: “in villam comites per Castrum Herculis transmittit Ulpiana.” The name is used in both its singular and plural form, Ulpianum, Ulpiana. Cf. Schol. ad Ptolem. iii. 9,6; “τὸ Οὐλπιανὸν, Οὐλπιάνα καλούμενον παρὰ τοῖς μεταγενεστέροις.” (Closs, ad loc.) The mention of Castrum Herculis, the Ad Herculem of the Tabula, the first station on the line Naissus-Ulpiana, fixes the route followed.

^{page 63 note b} Marcellinus Comes, Chron. sub anno, 518. See p. 89.

^{page 63 note c} Procopius, De Æd. iv. 1.: “ἧν δέ τις ἐν Δαρδάνοις ἐκ παλαιοῦ πόλις ἥπερ Ούλπιᾶνα ὠνόμαστο; ταύτης τὸν περίβολον καθελὼν ἐκ τοῦ ἐπιπλεῖστου (ἧν γὰρ σϕαλεϱὸς ές τὰ μάλιστα καὶ ὅλως ἀχρεῖος) ἄλλα τε ἀυτῇ παμπληθῆ ἑγκαλλωπίσματα ποιησάμενος, ἔς τε τὴν νῦν μεταθέμενος εὐκοσμίαν, σεκούνδαν αὐτὴν Ἰουστινιανὴν ἐπωνόμασεν. σεκούυδαν γὰϱ τῆν δευτέραν Λατῖνοι λέγουσι He built another city near it which he named Justinopolis, in honour of his uncle Justinus, an indirect piece of evidence that Procopius is right in making Justinian's fatherland Dardania. (See p. 137.)

^{page 63 note d} Acta SS. t. 35, p. 522. The Martyrium chiefly followed in the Acta SS. is headed: “Auctore Laurentio Monacho Rutiensi in Calabria,” and is written in Greek. The chronology is obscure, the account being divided between the reigns of Hadrian and Licinius!

^{page 63 note e} τὴν λιθοξόων ἐκπαιδευόνται τέχνην.. They had been originally in Constantinople but afterwards practised their craft at Ulpiana.

^{page 64 note a} Ἄρχων, ἡγεμών..

^{page 64 note b} See the chrysobull of Basil II. reorganising the Bulgarian Church (1020). Jireček, Gesch. d. Bulgaren, p. 202.

^{page 66 note a} Cf. Jireček, Die Heertrasse von Belgrad nach Constantinopel, p. 23.

^{page 67 note a} In Krasnichi is a ruin known as Giutet (Rouman, Civtat, Civetate = Latin, Civitas), but the Latin word is used in North Albania to signify any ruined castle.

^{page 67 note b} Drumi = SI. Drum = Byz. δρόμος..

^{page 68 note a} Podatci za istoriju Srpske Crkve (Contributions to the History of the Serbian Church), Belgrade, 1879, p. 65. M. Jastrebov informed me that he believed Roman remains to exist at Suharjeka, on the present route from Prisren to Lipljan. He had not, however, discovered any traces of a Roman line of way taking this route.

^{page 68 note b} At Dečani itself I could find no Roman monuments.

^{page 68 note c} Absurdly described as “Roman “by Isambert.

^{page 69 note a} See succeeding paper.

^{page 69 note b} The discovery of an inscription on the Kossovo Polje referring to this same legion (see p. 58 note^a), now adds additional probability to this conclusion.

^{page 69 note c} Putopis Stare Srbije (Travels in Old Serbia), p. 166.

^{page 69 note d} Milojević only copied the three that appeared to him most perfect.

^{page 69 note e} 1. D. M/VELS SADRAGI/TA MAG ․ DOM ․ VIX / AN XXXII ET SV / FIL BLAZZIZA V ․ / ․ ․ ․ PROCVL ․ VIX ․ ․ ․ ․ 2. MARCIVS FLAV/ET IVLIVS SER/G ․ VIX ANN XXII. 3. MERCVL ․ HAVE ․ BENEV/ALEA ․ ․ ․ / SALVTAS / D ․ M / MILIZZA BOSSINA / VIXIT ANNIS XXII YIPI/S ․ ․ ․ ․ CIA VIX ANNIS XXV/VLPIVS VVLCASSINVS / VIX ․ ANN ․ XXX / FLISB MER ․ ․ ․ VIVOS / F ․ C․ For the formula with which No. 3 begins compare that on the inscription from the Kossovo Polje (p. 58), VLP IONICE HAVE BENE VALEAS QVI ME SALVTAS.

^{page 70 note a} The temperature is only 76° Fahr.

^{page 70 note b} See Hajdeu, Resturile unei carti de donatiune depe la annul 1348, emanata de la Imperatul Serbesc Dušan, &c. (in Archiva istorica a Romaniei, Bucuresci, 1867).

^{page 71 note a} I have given some account of Valbona and the Rouman traces to be found in that part of the North Albanian Alps in a letter to the Pall Mall Gazette, “From the North Albanian Alps “(Sept. 14, 1880). In the map appended to this communication the upper Valley of the Valbona is for the first time given with approximate accuracy. In the last edition of the Austrian Stabskarte its place is occupied by a huge mountain mass.

^{page 71 note b} Copillu is said to be derived from the Latin pupillus, on the analogy of poturnichia from coturnicula.

^{page 71 note c} Presbyter Diocleas., Regnum Slavorum (Lucius, p. 288.)

^{page 71 note d} See p. 24.

^{page 71 note e} These local traces of Albanian and Rouman juxta-position, and the deductions at which I had quite independently arrived on linguistic grounds, entirely agree with the general results arrived at by Cihac in his analysis of the Rouman language. (Dictionnaire d'étymologie Daco-romane, préf. p. xiii.): “Le point capital et le plus important qui nous permet de juger des relations entre Roumains et Albanais dans le passé,—relations qui doivent avoir été des plus intimes,—sont les éléments concernant la langue que l'albanais possède de commun avec le roumain. Dans mes éléments latins de la langue roumaine et dans l'ouvrage présent, j'ai indiqué environ 500 mots latins, 1,000 mots slaves, 300 mots turcs, 280 mots grecs-moderne et 20 à 25 mots magyars pour l'albanais qui sont identiques avec les vocables correspondants roumains. Cette circonstance, assurément très-remarquable, ne peut être nullement fortuite, surtout en ce qui concerne les éléments latins qui ont subi dans les deux langues un changement d'acception presque analogue.” It is precisely this last circumstance that excludes Hajdeu's hypothesis that the community between the two languages is to be referred to an original relationship between the Illyrian and old Dacian languages.

^{page 72 note a} See p. 160.

^{page 73 note a} This monument has been described by Henzen in Eph. Ep. ii. p. 330, “ad ectypon quod misit Morten Noe.” My copy, however, which I made and very carefully collated on the spot, differs in line 9 and in other details. This monument, as well as the milestone (fig. 40), has been lately removed to the garden of the railway engineer at Kačanik; this place lying on the new line from Salonica to Mitrovica.

^{page 74 note a} Clemens and Priscus do not appear together in the Fasti Consulares. In 195 A.D. we find Tertullus and Clemens Consuls; in 196 Dexter and Priscus; it is probable, therefore, that the inscription belongs to one or the other of these years. Since this paper was communicated to the Society of Antiquaries a copy of this and the milestone on p. 74 has appeared in the Archäologisch Epigraphische Mittheilungen aus Oesterreich, 1883, part i. p. 145, on the strength of somewhat imperfect paper-casts sent by Signor Paolo Orsi of Rovereto. The name is there wrongly given ANDENVS and cos is added after PRISCO, which I did not see on the stone. With regard to the date Dr. Otto Hirschfeld remarks: “Vielleicht von J. 73 ? Der Name des Collegen im ersten Consulat des M. Arrecinus Clemens ist nicht bekannt.” But from the character of the letters the inscription cannot be of earlier date than the end of the second century of our æra. Sig. Orsi's copy of the milestone of Æmilian is still more imperfect, the important part being omitted.

^{page 76 note a} Aur. Victor, Epitome, c. xxxi; Eutropius, ix. 5; Zozimus, lib. i. speaks of Æmilian as Παιωνικῶν ἡγούμενος τάξεων = Dux Pannonicorum ordinum, and mentions a great victory gained by him over the barbarians who were then overrunning Illyricum.

^{page 76 note b} Forms like ABVERTO show the possibility of AB before v which was pronounced as w. AB VLCinio is a possible but not probable alternative.

^{page 76 note c} Some account of the antiquities of Viminacium has been given by Kanitz, Beitrage zur Alterthumskunde der Serbischen Donau, in Mitth. d. k. k. Central Commission, 1867, p. 28 seqq.) It was Trajan's chief base of operations in his Dacian campaigns, and was one of the principal stations of the Danubian fleet, as well as the headquarters of the Seventh Legion. The Leg. VII. Claudia is referred to on its autonomous coins and monuments, and tiles are found here with its stamp.

^{page 76 note d} See p. 153 seqq.

^{page 77 note a} In the Itinerary of Antonine, 118, M.P.

^{page 78 note a} A copy of this inscription has been given by Henzen in the Ephemeris Epigraphica, from a paper-cast that had been sent him by an engineer.

^{page 83 note a} A brief Account of Some Travels in Hungaria, Servia, Bulgaria, Macedonia, &c. by Edward Brown, M.D. of the College of London, Fellow of the Royal Society, and Physician in Ordinary to his Majesty. London 1673.

^{page 83 note b} See p. 102. The SHIANC of Dr. Edward Brown is evidently derived from the TRAIANO of the stone.

^{page 83 note c} Turquie d'Europe, T. 2, p. 354.

^{page 83 note d} Ephemeris Epigraphica, vol. ii. 498.

^{page 83 note e} Dr. Kenner Inschriften aus der Vardarschlucht, Sitzungsberichte der k. Akademie der Wissen. schapten, 1875, p. 276.

^{page 84 note a} Zur Kunde der Hämus Halbinsel. (Sitzungsberichte der K. Akademie der Wissenschaften, Wien 1881. H. 2, p. 437–499.) Prof. Tomaschek proposed to seek the site of Scupi near Leskovac in Serbia. Skopia he places in Pæonia.

^{page 88 note a} Literally “Bitter Water,” a common name for mineral sources throughout the South-Slavonic countries. The temperature was 75° Fahr.

^{page 89 note a} Comes Marcellinus (Ad. Ann. 518). “In Provincia Dardania assidno terræmotu xxiv. Castella uno momento collapsa sunt. Quorum duo suis cum habitatoribus demersa, quatuor dimidia ædificiorum suorum hominumque amissa parte destructa, undecim tertia domorum totidemque populi clade dejecta, septem quarta tectorum suorum tantaque plebis parte depressa, vicina vero (al. “vicinarum”) metu ruinarum despecta sunt. Scupus namque Metropolis, licet sine civium suorum hostem fugientium clade, funditus tamen corruit. Uno in Castello, regionis Canisæ, quod Sarnunto dicitur, ruptis tune terra venis et ad instar torridæ fornacis exæstuans diutinum altrinsecus ferventemque imbrem evomuit. Plurimæ totius Provinciæ montes hoc terræmotu scissi sunt, saxa que suis evulsa compagibus, devolutaque arborum (? Devolutæque arbores) crepido per xxx. passuum millia patens et in xii. pedum latitudinem deliiscens profundum aliquantis voraginem civibus castellorum saxorumque ruinas vel adhuc hostium incursiones fugientibus jussa* paravit.” The last paragraph is evidently corrupt, but the general sense is clear. Crepido here = fissura (Cf. Du Cange, s. v.). With this Dardanian “Sarnunto” I will venture to connect the Sarnoates, referred to on the Illyrian coins reading SAPNOATWN, and the Σαρνοῦς of Stephanus of Byzantium and Polyænus. I will even go further and suggest the emendation of the unknown (Βαρνοῦς) “Βαρνοῦντα” of Strabo (7, 7, 4), mentioned as lying on or near the Egnatian Way between Lychnidus (Ochrida) and Heraclea Lyncestis, into Σαρνοῦντα, and its identification in turn with the Σαρνοῦς of the coins, and the “Sarnunto”of Marcellinus. This attribution would bring down a corner of sixth century Dardania to the neighbourhood of Monastir, but it is not at least inconsistent with Procopius’ description of the “European Dardanians” as living above Dyrrhachium. The town and region of Monastir itself (at or near the site of the ancient Heraclea Lyncestis) was known in Byzantine times as Pelagonia, and we have here, I venture to think, a clue to the whereabouts of the Pelagia of a series of Illyrian coins that in all respects are companion pieces to those reading SAPNOATWN. On the other hand, the superior workmanship and Zacynthian affinities of the kindred Damastian coins would lead us to seek for the site of Damastion nearer the Epirote littoral. See p. 38.

^{page 91 note a} Gornjanska Crkva.

^{page 94 note a} Géographie d'Edrisi, traduite d'Arabe en Francais par P. Amédée Jaubert, t. ii. p. 289, 290.

^{page 94 note b} Edrisi describes Skopia itself as “a considerable town surrounded by many vineyards and cultivated fields.” From Skopia onwards he mentions a route to Kratova (Kortos), where two lines of communication branched, one to Nish, the other to Seres, Drama, and Christopolis.

^{page 94 note c} There is an apparent discrepancy in Edrisi's account. On p. 289 “Bolghoura “is mentioned as “a pretty town on the top of a high mountain,” four days from Scopia: on p. 290 “Boulghar” is mentioned as one day distant from Skopia.

^{page 95 note a} It is remarkable that in 409 A.D. we find Pope Innocent addressing a letter “Martiano Episcopo Naresitano” in which he refers to the “Clerici Naresienses” as having been nominated by the heretic bishop Bonosus (of Serdica). Farlato, Illyricum Sacrum, remarks on this, “Naresitanam ecclesiam nuspiam invenies in ecclesiastica geographia,” and would read “Naissitanam”: but the parallel form “Naresienses” and the high improbability of such a corruption of a well-known name like that of Naissus militate against the suggestion. Here at least we have an “Ecclesia Naresitana or Naresiensis of Byzantine date and within a territorial sphere over which a heretic bishop of the Metropolis of Dacia Mediterranea may have usurped authority. Dardania, it must be remembered, was at this time one of the “Five Dacias”; and, though the Metropolitan of Scupi seems to have claimed precedence over the Metropolitan of Serdica (see p. 138), Bonosus may have succeeded for a while in turning the tables.

^{page 96 note a} Ducange, Familiœ Augustœ Byzantinœ, p. 178, and 202. (Paris, 1680).Google Scholar

^{page 96 note b} Near Novipazar. (See p. 54.)

^{page 96 note c} Kinnamos Hist. Lib. III.

^{page 97 note a} It is well, however, to mention that the upper course of the Treška as depicted on the Austrian Stabs-karte is entirely erroneous. No tributary runs into it near St. Nikola, and the river itself takes a long straight turn to the West above that monastery, instead of running, as represented, from the North. On my sketch-map I have corrected the geography of this district so far as my explorations enabled me.

^{page 99 note a} Incompletely given by Engelhardt, loc. cit.

^{page 100 note a} It must be observed, however, that the stone appears to read DFC. c and not DEC.

^{page 100 note b} In the accusative form Darde-ne. Von Hahn Albanesische Studien, p. 236, compares the ancient derivation of the kindred Mysian race from a tree called in their language Μυσός = the Old Greek Ὀξύη, and instances Hesiod's account of Zeus creating the third or brazen race of men from ash trees (ἐκ μελιᾶν).

^{page 102 note a} It was undoubtedly from this stone that Edward Brown derived his inscription SHIAXC. See p. 83. No. 1 has been given by Dr. Kenner in a but slightly variant from on Herr Lippich's authority. See Sitzungsber. d. Wiener Akad. v. 80, p. 274; Eph. Ep. vol. iv. p. 82.

^{page 102 note b} This block was so heavy that it took six men to lever it sufficiently for me to read the inscription. The Hassanbeg stone has since been removed to the Konak at Skopia.

^{page 103 note a} Some interesting remarks on the Slavonic Rusalje, Rusalije, &c, and their connexion with the Roman Rusalia will be found, in Miklosich, Die Rusalien (Sitzungsberichte der k. Akad, d. Wissensch vol. xlvi. p. 386 seqq.), and W. Tomaschek, Über Brumalia und Rosalia (Sitzungsberichte, &c. vol. v. p. 351 seqq.). For the Roman Rosalia, see especially F. M. Avellino, Oposcoli (t. iii. p. 247 seqq.). Amongst the Lithuanians there is a June feast called Rasos Švente, which Miklosich shows to be the same celebration and derived from Rosas. Several inscriptions recording the celebration of the Rosalia on old Thracian soil have been discovered by Heuzey (Le Panthéon des rochers de Philippes, in Mission de Macédoine, p. 152 seqq.). The Roman Rosalia, at least in later times, seem to have been specially associated with the cult of Flora (Cf. Ovid, Fasti, lib. v.)

^{page 104 note a} See Jireć, Lexicon, s. v. Dodola. A Dodola song is translated by Mr. Ralston in his Songs of the Russian People, p. 228. The derivation is obscure.

^{page 106 note b} The modern Greeks have the Dodola in the form of Πορπροῦνα which is simply derived from the nasalized old Slovene form of Preperuga. The Wallachian name is Papeluga. Compare also Prporuše and Prpac, alternative male forms of the “Dodola”among the Serbs of Dalmatia (Vuk Stefanović Lexicon s. v.). Prpa is a Serbian word for ashes mixed with water.

^{page 106 note c} See p. 47.

^{page 107 note a} See p. 71.

^{page 107 note b} Vlachs near Skopia are mentioned under the Bulgarian Czar Constantine (1258–1277). See Jirečcek, Geschichte der Bulgaren, p. 218.

^{page 108 note a} An extraordinary instance of such a cult at the village of Selci belonging to the Clementi tribe is given in Dečanski Prvenac, Novisad (Neusatz), 1852, p. 81.

^{page 108 note b} Tylor (Primitive Culture, vol. ii. p. 260, 2nd ed.), who cites Dalton, Kols in Tr. Ethn. Soc. vol. vi. p. 35.

^{page 108 note c} Roman de Rou, ii. 6399. (Ed. Andresen ii. 283)

“La fontaine de Berenton

Sort d'une part lez un perron;

Aler soleient ueneor

A Berenton par grant chalor,

E a lor cors l'eue espuisier

E le perron desus moillier,

Por co soleient pluie aueir;

Issi soleit iadis ploueir

En la forest e enuirun

Mais io ne sai par quel raison.”

Cf. Grimm. Deutsche Mythologie (4th Ed.) vol. iii. p. 494. At Kulen Vakup in Bosnia I came upon the reverse of this method. There, sacred stones are let down in a net into the spring to produce rain. If the stones were to drop out of the net a great flood would ensue. See my Illyrian Letters, p. 109. For another Breton parallel see Crestien de Troies, Li romans dou Chevalier au Lyon, v. 387, seqq.

^{page 111 note a} See p. 102.

^{page 111 note b} NEMESI / SANCTAE / CAMPESTRI. PRO SA / LVTE. DOMINORVM. / NN. AVGG. P.AEL.P F / AELIA PACATVS. / SCVPIS. QVOD. COH DOC / TOR. VOVERAT. NVNC / CAMPI. DOCTOR. COH. I. / PR. PV. SOMNIO. ADMO / NITVS. POSVIT. L. L. In Kellerman, Vigil. Rom. No. 119.

^{PAGE 113 NOTE a} A mutilated and blundered version of this inscription was communicated by “a Belgrade professor ”to M. Engelhardt and published by him in the Révue Archéologique, vol. xxvi. p. 137, from which it has been copied into the Ephemeris Epigraphica, vol. ii. p. 497. It is strange that there should have been any difficulty about this clear and beautifully-cut inscription.

^{page 115 note a} D.M / CAELIDIA. SE / CVNDA. VIX. AN L / H.S.E.CL / HERCVLANVS MA / RITVS B.M.P.. Given in Eph. Ep. vol. ii. 498.

^{page 115 note b} C.I.L. v. 5963, NVMMIA AVRHA, of Canusium; ix. 395, ATILIA AVRA, at Milan; x. 2438, MARCIA AVRA, at Naples.

^{page 115 note c} See Mommsen, Eph. Ep. loc. cit.; and of. C.I.L. v. 1892, where in the case of the ornamenta duoviralia he observes: “Ornamenta duoviralia cum non soleant concedi vivo nisi ei qui per legem duovir fieri non possit, crediderim et hie et in aliis similibus (ut Henzen 7172), ubi ingenuis en tribuuntur, significari ornamenta post mortem decreta, sepulturæ; causa.”

^{page 115 note d} C.I.L. iii. 649, 659.

^{page 116 note a} Compare for the abbreviated character of the lines the almost contemporary inscription on a six-sided base of a statue of Marsyas erected PEO SA / LVTE / ET IN / COLV / MITA / TE D D / N N VA / LERIA / NIET/GALLI/ENI/AVGG &c. at Verecunda in the Province of Numidia (C.I.L. viii. 4219). The whole inscription in this latter case extended over three sides of the base containing severally twelve, fourteen, and eight lines.

^{page 116 note b} Sextus Rufus, in Brev. “Dacia Gallieno imperatore amissa est.” For Aurelian's Dacia cf. Fl. Vopiscus, 39, from whom Eutropius (ix. 15) copies. Mœ;sia is described as “deperdita ”at this time.

^{page 116 note c} “Fusco (lege Tusco) et Basso Consulibus” the date of Ingenuus' revolt (Treb. Poll, XXX Tyranni. 8), which was caused by the imminence of this Sarmatian invasion.

^{page 116 note d} Treb. Pollio. Triginta Tyranni ix. “Claudius Regilliano (sic) multam salutem. Felicem Rempublicam quæ; te talem virum habere rei castrensis bellis his meruit, felicem Gallienum, etiamsi ei vera nemo nee de bonis, nee de malis nuntiat. Pertulerunt ad me Bonitus et Celsus stipatores Principis nostri qualis apud Scupos in pugnando fueris quot uno die præ;lia et qua celeritate confeceris. Dignus eras triumpho si antiqua tempora exstarent. Sed quid multa ? Memor cujusdam ominis cautius velim vincas. Arcus Sarmaticos et duo saga ad me velim mittas, sed fibulatoria, cum ipse miserim de nostris.” The “omen” referred to was no doubt the fate of Ingenuus.

^{page 117 note a} Treb. Pollio. Triginta Tyranni, viii. “In omnes Mœ;siacos, tam milites quam cives, asperrime sæ;viit, nec quemquam suæ crudelitatis exsortem reliquit: usque adeo asper et traculentus ut plerasque civitates vacuas a virili sexu relinqueret.”

^{page 117 note b} Ib. “Perimendus est omnis sexus virilis, si et senes atque impuberes sine reprehensione nostra occidi possent. Occidendus est quicumque male voluit, occidendus est quicumque male dixit contra me, contra Valeriani filium, contra tot principum patrem et fratrem. Ingenuus factus est imperator. Lacera, occide, concide.”

^{page 117 note c} Treb. Poll. Triginta Tyranni, ix. “Gentis Daciæ;, Decebali ipsius ut fertur affinis.”

^{page 117 note d} Cf. Lactantius de Mortibus Persecutorum C. xxvii. “Olim quidem ille, ut nomen Imperatoris acceperat, hostem se Romani nominis erat professus, cujus titulum immutari volebat ut non Romanum imperium sed Daciscum cognominaretur.”

^{page 118 note a} The revolt of Regalianus appears to have taken place about the date of Gallienus' Decennalia, A.D. 263. Cf. Clinton Fasti Romani; ad annum.

^{page 118 note b} C.I.L. vi. 1106.

^{page 118 note c} The language of the present inscription recalls the lines of Calpurnius (Ecl. IV.)

“In uno

Et Martis vultus et Apollinis esse notatur.”

The flattering comparison of Calpurnius is, however, addressed, as Moriz Haupt has conclusively shown (De Carminibus bucolicis Calpurnii et Nemesiani), to Nero and not, as earlier commentators supposed, to Carinus or Gallienus himself.

^{page 118 note d} Treb. Pollio. Duo Gallieni. “Fuit enim Gallienus (quod negari non potest) oratione, poëmate, atque omnibus artibus clarus. Hujus est illud epithalamium quod inter centum poëtas præ;cipuum fuit.”

^{page 119 note a} Thus we find the Præ;fectus of the 5th Macedonian Legion at Potaissa in Dacia erecting a votive altar to Azizus “Bonus Puer Conservator”for the health of Valerian and Gallienus. C.I.L III. 875. Julian Or. IV. mentions the worship of Azizus at Edessa in conjunction with that of the Sun, and notices that Jamblichus identifies this god with Ares. From inscriptions found at Apulum, however, as Mommsen has pointed out, Azizus is seen to be the equivalent of Apollo Pythius. See C.I.L. III. 1133.

^{page 119 note b} Treb. Pollio. Gallieni Duo: “Statuam sibi majorem colosso fieri præ;cepit, Solis habitu, sed ea imperfecta periit … Poni autem illam voluerat in summo Esquiliarum monte, ita ut hastam teneret, per cujus caput infans ad summum posset ascendere. Sed et Claudio et Aureliano deinceps stulta res visa est, &c.”

^{page 119 note c} In my father's cabinet: unpublished.

^{page 120 note a} Cf. C.I.L. vi. 2385.

^{page 120 note b} Ami Boué, Turquie d'Europe, 2, 354; C.I.L. iii. 1696; PRO SALVte imp. caes. L. septimi severi pertinacis Aug. Arab. / ADIAB. POnt. Max.… / M. AVEELI Antonini caes.…. The A of ADlABenici is clear.

^{page 120 note c} Two from the neighbouring Markova Rjeka district (Figs. 43, 46).

^{page 122 note a} A copy of the inscription sent by the Austrian Consul Lippich was published by Dr. Friedrich Kenner (Sitzungsberichte d. k. Ahademie d. Wissensch. vol. 80, p. 275, and see Eph. Ep. vol. iv.), but the relief is inaccurately described. In Dr. Kenner's version, line 6, OBVLCIIA.

^{page 122 note b} See Henzen, Sui militi peregrini e frumentarii, in Bullettino dell' instituto di Corr. Archeologica, 1851, p. 113 seqq.

^{page 122 note c} Aurelius Victor, De Cœ;ss. 39, speaking of Diocletian, says: remoto pestilenti frumentariorum genere quorum nunc agentes in rebus simillimi sunt.

^{page 122 note d} C.I.L. iii. D. xx. xxii. xxxiv.

^{page 124 note a} Cf. Forcellini Lexicon (Ed. De Vit), s. v. Præ;ceptio. “Per præ;ceptionen dare, legare, relinquere, est ita dare ut percipiatur ante quam tota hereditas dividatur et partes aliis coheredibus distribuantur.” Julian, Dig. 30, 122, “Si heres centum præ;cipere jussus sit.”

^{page 124 note b} In 184 Ælianus had been consul in conjunction with Marullus. The name of Crispinus however squares better with the letter-space at our disposal, which has been very accurately observed throughout this inscription.

^{page 124 note c} Cf. especially a bas-relief of the Genius of Viminacium represented as a stoled female figure with her right hand on the bull of the 7th, Claudian, Legion, and her left on the lion, which here stands for the 4th Legion (figured by Kanitz, Beitrage zur Alterthumskunde der serbischen Donau, in Mitth. d. Central. Comm. 1867, 28 seqq.) The same device is common on the coins of this city.

^{page 124 note d} Lib. iv. C. 23: “Kαi ἓβδομοι οἱ ἐν τῇ Μυσίᾳ τῇ ἄνω οἱ τὰ μάλιστα Κλαυδίειοι ὠνομάζαται.”

^{page 124 note e} “Præ;fecturæ Leg. vii. Claudiæ Viminacio.”

^{page 124 note f} Mommsen, C.I.L. iii. 272. Cf. Inscriptions at Narona (1813, 1814, 1818), Salona (2014, 2019, 2040, 2033, 2048, 2071), at Tilurium (Gardun), (2709, 2710, 2714, 2716, 2717), where Mommsen fixes their Prætorium, at Nedinum (2882), and at Jader (2908, 2913). Detachments of this Legion are found serving in Syria and Asia.

^{page 125 note a} C. I. L. iii. 1676.

^{page 125 note b} Bithynian, C. I. G. 3795; cf. Tomaschek, Brumalia, &c. p. 386, for this and other instances. Tomaschek also compares names like Rascupolis, Abropolis.

^{page 125 note c} The name occurs in Dalmatia, Italy, and other parts of the Empire.

^{page 125 note d} Revue Archéologique, VI. Année (1865), p. 451. Tomaschek, op. cit. p. 392, cites other instances, I. R. N. 513, EVTICHIA QAE ET BVTIN; 2810, C. RAVONIVS. CELER. QVI. ET. BATO. SCENOBARBI. NATIONE. MAEZeius (Dalmatian), &c.

^{page 126 note a} I was informed that some had been lately thus transported to Salonica from Zlokucani. Others have in the same way been removed by the Turkish authorities from Bardovce. Monuments with sculpture are more especially sought for by the Turkish authorities as they are thought to have a monetary value. No pains are taken in such cases to preserve a record of the locality where the monuments were found.

^{page 127 note a} Cf. Tomaschek, Zur Kunde der Hœ;mus-Halbinsel (Sitzungsb. d. k. Akad. d. W. 1881. H. 2, p. 446.) page 127 note b A Dardanian with the Illyrian name Epicadus is mentioned on an inscription at Rome C.I.L. VI. 2845.

^{page 127 note c} Cf. also the uncertain fragment from Taor (p. 145) and the later Byzantine inscription on the walls of Skopia.

^{page 128 note a} Less the part which was now incorporated in Dacia Mediterranea. Naissus itself had been included in the older and more extensive Dardania by Ptolemy.

^{page 128 note b} “Paregorius a Dardania de Scupis ”: the other Dardanian Bishop who attended this council was Macedonius of Ulpiana. Mansi, Conc.

^{page 128 note c} Cod. Theod. De Palatinis 1. 2, dated “Scopis.”

^{page 128 note d} Cod. Theod. De Decurionibus 1. 119, dated “Scupis.’

^{page 128 note e} S. Paulini Nolensis C. xxx: De reditu Nicetœ Episcopi in Daciam: see p. 163 seqq.

^{page 132 note a} Not. Orientis IX. The Ulpianenses and Mer(i)enses are also mentioned; the names of which connect them with the Dardanian towns of Ulpiana and Merion.

^{page 133 note a} Excerpta e Malchi Historia. (Ed. Bonn, p. 255).

^{page 133 note b} “Johannes Episcopus Sacrosanctæ; Ecclesiæ; Scopinæ;, Metropolitanæ;.” Mansi. viii. 13.

^{page 133 note c} “Gelasius Episcopis per Dardaniam sive per Illyricum constitutis … Audientes orthodoxam vestræ dilectionis in Christo constantiam.” Mansi viii. 46.

^{page 133 note d} Mansi viii. 408.

^{page 133 note e} Marcellinus Comes, in Chron: Anastasius was constrained to send back the Bishops of Naissus and Pautalia, ob metum Illyriciani Catholici militis. Prof. Tomaschek rightly, I think, connects the Roman and Italian sympathies of the Illyrian church and army with the prevalence of the Latin tongue in the interior of the peninsula.

^{page 133 note f} In Chron. sub anno.

^{page 133 note g} See p. 89.

^{page 134 note a} De Æd.. iV. 1.ἐν Δαρδάνοις που τοῖς Εὐρωπαιοῖς,οἳ δὴ μετὰ τοὺς Ἐπιδαμνίων ὅρους ᾤκηνται, τοῦ ϕρουρίου ἄγχιστα ὃπερ Βεδεριανὰ ἐπικαλεῖται,χωρίον Ταυρήσιον ὄνομα ἧν, ἔνθεν Ιουστινιανὸς βασιλεὺς ὁ τῆς οἰκουμένης οἱκιστὴς ὥρμηται. Τοῦτο μὲν οὗν τὸ χωϱίον ἐν βϱαχεῖ τειχισάμενος κατὰ τὸ τετϱάγωνον σχῆμα καὶ γωνίᾳ ἑκάστῃ πύϱγον ἐνθέμενος Τετραπυργίαν εἶναί τε καὶ καλεῖσθαι πεποιήκε.Παϱ᾽ αὐτὸ δὲ μάλιστα τὸ χωρίον πόλιν ἐπιϕανεστάτην ἐδείματο, ἥνπερἸουστινιανήν ὠνόμασε πϱίμαν (πϱώτη δὲ τοῦτο τῇ Λατίνων ϕωνῇ δύναται)ταῦτα τῇ θϱεψαμένῃ τροϕεῖα ἐκτίνων.”

^{page 135 note a} Novella Constit. ii. “Multis et variis modis nostram patriam augere cupientes, in qua primo Deus præ;stitit nobis ad hunc mundum, quem ipse condidit, venire, et circa Sacerdotalem censuram eam volumus maximis incrementis ampliare, ut Primæ; Justinianæ; patriæ; nostræ; pro tempore sacrosanctus Antistes non solum Metropolitanus sed etiam Archiepiscopus fiat, et cæ;teræ; provinciæ sub ejus sint auctoritate, id est tam ipsa Dacia Mediterranea quam Dacia ripensis necnon Mysia Secunda, Dardania et Præ;valitana Provincia et secunda Macedonia et pars secundæ; etiam Pannoniæ; quæ; in Bacensi est civitate” … necessarium duximus ipsam gloriosissimam Præ;feoturam, quæ in Pannonia erat, in nostra felicissima patria collocare cum nihil quidem magni distat a Dacia Mediterranea Secunda Pannonia.” So too in Nov. 131 Dacia is placed first amongst the provinces under the jurisdiction of the Archbishop of Justinian's father-land.

^{page 135 note b} As for example, when he speaks of the “European ”Dardanians, and of their living above the “Epidamnians.” The name of Epidamnos had long given way to that of Dyrrhachium.

^{page 135 note c} Ptol. Geogr.

^{page 136 note a} W. Tomaschek, Miscellen aus der alten Geographie in Zeitschrift für die Oesterreicliischen Gymnasien 1874, p. 659.

^{page 136 note b} Hermes, B. vi. p. 323 seqq.

^{page 136 note c} “Ἰουστῖνος ἐκ Βεδεριανοῦ ϕρουρίου πλησιάζοντος Ναἷσσῳ” op. cit. p. 339. Justin was assisting the Emperor Anastasius against the Isaurian rebels in the capacity of Hypostratêgos.

^{page 136 note d} “Die Sache ist entschieden.” As to the opinion—supported by weighty arguments by Mannert, Hahn, and Tozer—that Scupi and Justiniana Prima were identical, Prof. Tomaschek thinks it not worth the trouble of refuting. “Diese Meinung zu wiederlegen verlohnt sich nicht der Mühe.” Miscellen. &c. p. 658.

^{page 136 note e} Geographie der Griechen und Römer, vii. p. 105 (Landshut). Mannert, however, had not observed the difficulty raised by Justinian's attribution of this city to Dacia Mediterranea.

^{page 136 note f} See p. 143.

^{page 136 note g} “Gospodin”Konstantin, Lord of Northern Macedonia († 1394), well-known in Serbian epic as the friend of Marko Kraljevic. In 1500 the territory formerly held by him was still known as Zemlja Konstantinova. In 1559 his City of Velebužd or Banja (this latter name derived from its hot-baths) appears in an Italian Itinerario as “Constantin-bagno.” Küstendil is simply the Turkish form of Konstantin. See Jireček, Gesch. d. Bulgaren. p. 333.

^{page 137 note a} See p. 58.

^{page 137 note b} He was of course of Thracian descent.

^{page 138 note a} Mansi, x. 329. “De Paulo Docleatinæ Civitatis episcopo lapso.” Justiniana Prima seems to be thus brought into a certain geographical connexion with Scodra (Scutari d'Albania), from which place as we have seen a line of Roman road led to the Dardanian City of Ulpiana (Justiniana II.), and thence to Scupi.

^{page 138 note b} Marius Mercator, in Appendice ad Contradictionem 12 Anathetismi Nestoriani, “Sardicensis Bonosus qui a Damaso urbis Romæ; episcopo præ;damnatus fuit:” Le Quien; Oriens. Christianus, t. ii. p. 302. Farlato III. Sac, t. viii. p. 34, endeavours to make Bonosus Bishop of Naissus, but on no valid grounds. His statement would anyhow not affect the present argument, as Naissus was also in Dacia Mediterranea.

^{page 138 note c} Mansi, ix. p. 199.

^{page 138 note d} Illyricum Sacrum, t. viii. p. 17.

^{page 138 note e} Oriens. Christianus, t. ii. p. 310.

^{page 139 note a} See Mommsen, Révue Archéologique, N. S. xiv. p. 387. The words of Rufus in describing the formation of Aurelian's Dacia are: “Per Aurelianum, translatis exinde Romanis, duœ Daciœ in regionibus Moesiœ et Dardaniœ factœ sunt.”

^{page 139 note b} Not. Or. iii. 14.

^{page 140 note a} D. B. Goth. ii. pp. 80, 418 (Bonn ed.).

^{page 140 note b} Compare Jornandes' Sirmis, &c. In Ravennas the form Scupis occurs, cf. Londinis, &c.

^{page 140 note c} Hist. vii. 2 (Bonn ed. p. 272). Τὰ γὰρ Ζαλδαπὰ καὶ Ἄκυς καὶ Σκόπις καταπρονομεύσαντες, &c.

^{page 140 note d} Acta S. Demetrii, c. ii. It is there mentioned as a chief cause of the second Slavonic onslaught on Thessalonica that that city sheltered escaped “mancipia” from the interior of Illyricum. One city only ought not to be allowed to hold out when all the other cities and provinces round had been made void of Roman habitation; “hæc autem” (to quote the Latin version) “sola superesset omnesque e Danubii partibus Paunoniaque et Dacia et Dardania reliquisque provinciis et urbibus transfugas reciperet atque in sinu suo foveret.” The citizens of Naissus and Serdica are specially mentioned.

^{page 141 note a} Nikephoros Bryennios, iv. 18 (Bonn ed. p. 148), in the eleventh century still calls Skopia by its ancient name of Σκουποι and places it correctly on the Vardar as he tells us the Axios was then called.

^{page 141 note b} Skylitzes and his copyist Kedrenos (Bonn ed. ii. 527). The revolted Bulgarian Prince, Peter Deljan, marches “διὰ τε Ναϊσσοῦ καὶ τῶν Σκουπίων τῆς Μητροπόλεως Βουλγαρίας” (A.D. 1040). When Basil organised the Bulgarian Church in 1020 the Bishop of Skopia was assigned 40 Kleriki and 40 Πάροικοι, putting it on a level with the largest Bulgarian Sees (see Jireček, Geschichte der Bulgaren, p. 202).

^{page 141 note c} By the neighbouring Albanian tribes, the best local representatives of the Roman provincials, the town is still called “Scup.”

^{page 141 note d} Reise von Belgrad nach Salonik, p. 156. Tauresium might easily represent a Vicus Taurensium, pointing to some form with which Taor would connect itself. Neither Taor nor Bader appears to be of Slavonic origin. As a set-off to this, Prof. Tomaschek, who seeks his Justiniana Prima near Kuršumlje, has sought to connect the name of Tauresium with that of the village of Tovrljan in the Toplica district.

^{page 142 note a} Op. cit. p. 157. “Hier war kein Platz für Prokop's Tetrapyrgion, doch erzählten die Bauern, dass sie beim Beackern der auf der Platte oberhalb der Dorfes gelegenen Felder auf Cementsubstructionen stiessen, und bejahten unsere Frage, ob diese ein Viereck bildeten, doeh möchten wir durch diese Bejahung die Frage noch nicht als unwiderruflich entschieden betrachten. Die auf der Platte lagernde Schneedecke machte die Untersuchung derselben durch den Augenschein unmöglich.” The peasants also spoke of a quadrangular tile conduit leading to these remains.

^{page 142 note b} Op. cit. p. 158.

^{page 142 note c} Op. cit. p. 162.

^{page 142 note d} See p. 110.

^{page 143 note a} I ․ O ․ M ․ / IVNONI ․ REG ․ MIN / SANC ․ CETERIS QVE / DIIS DEABVSQVE / OMNIBVS ․ SACR M / AVR ․ TITIANVS ․ BF ․ / COS ․ LEG ․ VII ․ CL ․ / V ․ S ․ L ․ M

^{page 143 note b} A neighbouring village.

^{page 144 note a} The translation of the inscription as given to Hahn (p. 162) was of a curious kind: “die Inschrift, … wenn mann sie uns richtig übersetzt hat, den Arzt eines türkischen Pascha's, welcher dessen Gattin von der Unfruchtbarkeit heilte, als den Wiederhersteller des von Justinian gegründeten Klosters nennt “(!)

^{page 144 note b} Op. cit. p. 158: “Leider stand das Postament auf dem Kopfe und ist die Inschrift so verwischt dass wir nur mit grosser Mühe einige roh gearbeitete slavische Charaktere erkennen konnten.”

^{page 145 note a} The foundations about the corner A were very indistinct, and in order to ascertain the outline of this part of the castellum excavations would be necessary. The measurements given are approximate.

^{page 146 note a} “Selo Tavor, gradište…. s jezerom Jelatnim.” (Šǎfaříik. Památky 25; quoted in Jireček Geschichte der Bulgaren, p. 79.

^{page 146 note b} Given in Hahn: Albanesische Studien, p. 122. When I saw this inscription, it was broken into two fragments and used as a support for the wooden post of a verandah in the Turkish Governor's Konak.

^{page 146 note c} Hahn. Drin und Vardar-Reise, p. 115. The name of the prelate in whose honour the inscription (of colossal size) was put up has disappeared, but we are told:

“σκηνὴν ἐγείρας τὸν θεόϕανον νόμον ἒθνη τὰ Μυσῶν ἐκδιδάσκει πανσόϕως.”

^{page 147 note a} The beginnings of this form of construction may be traced in the walls of the Imperial Palace at Trier.

^{page 147 note b} This Han has been well-described by Mr. Tozer, Highlands of Turkey, vol. i. p. 367. The Sulei Han is another edifice of considerable antiquity. In the Fererli Han are said to be concealed inscriptions. These buildings at present afford lodgings and warehouses for merchants. On the piers of the Kurshumli Han many names of old Ragusan merchants are to be seen painted in red letters, e. g. “ANNO DOMINI 1777 MARINVS ZAMAGNA POST BREVEM MORAM ․ ․ ․ ․ ․ ․ ․” I also noticed the names of Lucich and Radegla. On the outside wall of the Han is a Turkish inscription.

^{page 148 note a} The Turks attribute the construction of the Kuršumli Han to a certain Mahmoud Pasha.

^{page 148 note b} P. 101.

^{page 149 note a} For pointed arches in the Aqueduct of Segovia, built in Trajan's time, see Archaeologia, vol. IV. page 410, note.

^{page 150 note a} An older Christian tradition regarding the aqueduct is, however, mentioned in the relation of the Ragusan ambassadors who passed through Üsküp in 1792. “Nella vicinanza di Uschiup videro un antico acquedotto mezzo rovinato volgarmente detto Gerina Ciupria, cioé Ponte di Jerina moglie di Giorgio Despot, per che da lei fabricato acquedotto fatto a forza di archi molto simile a quello di Pisa.” Jerina or Irene, wife of the Serbian despot George Brankovich, is popularly credited with many buildings throughout those countries. The description “mezzo rovinato”is interesting as showing that some restoration of the work must have taken place since the end of the last century.

^{page 153 note a} See p. 65 seqq.

^{page 154 note a} This stone had been previously observed by Von Hahn (Reise von Belgrad nach Salonik, 239. C. I. L. iii. 1697). His observations were conducted, however, under most unfavourable climatic conditions, and his copy is inaccurate in every single line. He made out the dedication to be one to Mithra.

^{page 155 note a} About two hours distant from Kumanovo to the East, at Nagurié, is a splendid example of an old Serbian church, with an inscription recording its erection by King Miljutin, and frescoes within of the King and his consort Simonida. Like Dečani, it is evidently the work of a Dalmatian architect, and represents a compromise between Italian and Byzantine styles. I must however reserve its description for another occasion.

^{page 156 note a} The coins included silver pieces of the Pæonian princes, Patraos and Audoleon, Macedonian, Roman, and Byzantine. Pæonian coins seem, more abundant in this district East of the Karadagh than in the immediate environs of Skopia. They are also abundant about Vranja in the upper valley of the Bulgarian Morava.

^{page 156 note b} Mustafa had picked up a little Italian from some workmen engaged on the new Serbian line. Amongst words in the local dialect which struck him as like Italian he instanced Szavle=Sand. (Cf. Ital. Sabbia, Rouman, Sablu), Plop or Plep=poplar (Ital. Pioppo, Macedo-Rouman Plop), Sielce=willow (Italian Salice, Macedo-Rouman Salice or Salce), Supra=above (It. Sopra, Rouman Supra, ordinary Albanian Siper), Ca'olli also Cavolli, horse (It. Gavallo, Rouman, Callu, ordinary Albanian Colli or Calli), &c.

^{page 160 note a} Since this account was written, I see that the ruins of Zlata are alluded to by Von Hahn (Reise von Belgrad nach Salonik, p. 55). On his way from.žitni Potok to Leskovac, he passed the ruins of “Slata”—the Albanian form of the Serb Zlata. He saw upon the hill the remains of an “Umfassungs-mauer” of hard burnt brick and firm cement, and speaks of the remains of a bridge on both sides of the brook, by which he certainly refers to the dam. Hahn apparently had no opportunity to explore the remains further, but he noticed their Roman appearance and rightly brought them into connection with the Roman road from Naissus to Ulpiana. He learnt from an Arnaout Aga a local tradition that Sultan Murad had taken the stronghold from a certain “Kralica” (Queen).

^{page 161 note a} Donau-Bulgarien und der Balkan, Bd. 1, p. 157 seqq. (1875).

^{page 161 note b} See Fragments inédits de l'historien grec Priscus recueillis et publiés par C. Wescher in Revue Archéologique N.S. vol. xviii. (1868)Google Scholar, p. 86 seqq. Cf. Jireček, Heerstrasse, p. 21. Priscus, however, erroneously calls the river “the Danube.”

^{page 163 note a} The Turkish Mustafa Pasha Palanka.

^{page 163 note b} The part of the stone containing the name of Carus is broken off: the R.. I (the last letter doubtful) after Caeino is enigmatical. To restore REGI would be too bold, though we recall Vopiscus' curious statement with regard to this Emperor “Regem denique illum Illyrici plerique vocitarunt” (Vop. Carinus).

^{page 163 note c} S. Paulini Nolensis, c. xxx, De Nicetœ reditu in Daciam, written about the year 398.

^{page 163 note d} i. e. the Provincials of Dacia Mediterranea and Dacia Ripensis. Remesiana itself was in Dacia Mediterranea.

^{page 164 note a} C. I. L. III. p. 268 (No. 1685, 1686). This site, as Mommsen justly observes, must not be confounded with that of the Dardanian Ulpiana.

^{page 164 note b} See Archaeologia, vol. XLVIII. p. 86–7.

^{page 165 note a} DEVS NOSTER ✝ PROPITIVS ESTO REI PVBLICAE ROMANAE.

^{page 165 note b} Rossi, De:—(Bullettino di Archeologia Cristiana, 1874Google Scholar, p. 145), seqq., where see also the dedication of S. Pietro in Vincoli.

^{page 165 note c} Op. cit. 1878, p. 37.

^{page 166 note a} Op. cit. 1872, p. 14. The Legates of the Apostolic See in the East wrote to Pope Hormisdas in 519, that Justinian, then Comes,—“basilicam Sanctorum Apostolorum (Petriet Pauli) constituit, in qua desiderat et beati Laurentii Martyris reliquias esse,” &c.

^{page 167 note a} Here Philippœos is to be taken not as referring to Philippi, but as an epitheton ornans for Macedonia in general. Thessalonica was the special city referred to.

^{page 167 note b} Accepting Pagius' admirable emendation, “Stobitanam” for “Tomitanam.” Tomi lay far away from any possible line of route that St. Nicetas could have taken.