The present article shows, for the first time publicly, some of the results I reached after more than 20 years of study upon Procope's "De Aedificiis". Some of the scholars might find my ideas too bold and too risky and might judge that I intervene too often in the text to consider them worthy to be given credit to. I was already told that, but I can't agree. Even if it is so, he who took his time, like me, to analize in its smallest details this text of such a tremendous importance for the Balkanic toponomastics, surely could not ignore the challenge of the so many and so obvious mistakes of all kind met in the, unfortunately, so few manuscripts of this work. I, of course, do not compare myself with the great Bulgarian scientist Veselin Beševliev, but he also, when trying to put an order in this huge material, felt obliged to suggest a lot of conjectures and emendations, some of them obvious and natural, others "bold", others anyway better than their actual wrong forms in the manuscripts. My conviction is that the confusion we find today in the two annexed lists of the descriptive text, which contain more names than in all other ancient geographical works put together, do not fall in Procope's responsability, but were made - in fully good faith - by those who later copied this text after a badly damaged archetype and this way maybe saved it from total destruction. I also think that many of these mistakes can be corrected and that the prize for that worths the pain.
* * *
In the lines to follow I take as basic reference the book of Veselin Beševliev Zur Deutung der Kastellnamen in Prokops Werk "De Aedificiis", Amsterdam, 1970 (abbreviated below as PKN). Being a short version of a section of my doctorat thesis, in this article I shall repeat nomore the information that my lector would easily find in this remarkable work of the Bulgarian scientist. Whenever you will meet Thracian names with no reference, they are to be found in Dimiter Detschew's Die thrakischen Sprachreste, Wien, 1957, sub voce. I ask those who take their time to read this text through, to share with me their impressions, their critic remarks and to add their own information (I assure them that their remarks will be exactly and properly quoted). After all, what interests us all is to reconstruct, based on our common knowledge, a lost past. Thank you.
* * *
Under the Title ᾽Εν Εὐρώπῃ(145,17)
145.18. Λυδικαί - Hapax. Etymology by Beševl. PKN p.127.
145.19.᾽Ελαίαι (ἐλᾶναι A). In Europe. It is ᾽Ελαιοῦς from Ptol. 3.11.9, Elaeus by Plin NH 4,49, fortress and promontory in the Thracian Chersonesus, province of Europe, see Beševl. PKN p.127.
Under the Title ᾽Εν ῾Ροδόπῃ τὰ καινούργια (145,20-21)
145.22. *Κασσήρα (κασεήρα V, κασέηρα A). Hapax. Possible denaturation of a *Κάσσηρα, with the graphical confusion Σ/Ε. We find such a form by Plin.NH 4,38: Cassera (considered as a wrong lection for Assara, town in Pallene — Dečev TSR p.31). But the alternance K-/zero appears in other surely correct macedonian names as well, such as in Καλίνδοια/᾽Αλίνδοια - Dečev TSR p.223, Κάσσανδρος/῎Ασσανδρος , Κάσσωπες/ Asopus etc Russu Mac.s.v). Other explanations by Beševl. PKN p.127. Therefore it could be in western Rhodope or eastern Macedonia.
145.23. Θεοδωρούπολις. The toponym also appears in the list of Haemimontus (147,40). Beševl. PKN p.72 and 127 quotes other attestations: in some lists of bishop sees (in Eparchia Europe)1 and in a protobulgarian inscription2, which permit to localize the town at Saray (European Turkey). Beševliev thinks it must have been near the intersection point of the borders of three provinces: Europe, Rhodope and Haemimontus, but his assertions cannot be sustained entirely. On the other hand, accepting that we deal here with an interpolation from the list of Europe3 , we are no more compelled to move this town to the west, so it may be very well placed on the border between Europe and Haemimontus, south from Tzoides, that means near Saray (s. also Dečev TSR p.498 s.v. Τζοΐδης).
145.24-25. *Τὸ τοῦ θράσους ([τ]ου) Δανέλου (τὸ τοῦ θράσου σουδάνελ A τὸ τοῦ θρασουσουδανελ᾽ Hapax. Beševl. PKN p.12-13 and 127 restores here two toponyms: τὸ τοῦ Θράσους "Kastell des Mutes" and Δανέλου – vulgar spelling of the biblical name – with the dittography of the diphthong ου. In this case the second name would be [τ]οῦ Δανέλου. But, once accepted the dittography, why not consider the rest as a single toponym: τὸ τοῦ θράσους<ου> Δανέλ(ου) "(e.g. praedium) audacis Daneli" (from θρασύς,-οῦς = audax). As, in this form, it looks to be the name of a farm (praedium) or a villa of an important character, there is a slight possibility that it would be situated closer to the capital, somewhere in Europe.
145.26. Μάνδεπα (μούνδεπα V). Hapax. Beševl. PKN p.127 admits that the original form must have had an -α- (but he doesn't correct) and that the second lexical formant is the Thracian -απα "river". For the first part we might consider a relation with the placename in Pallene Μένδη (etnicon Μενδαῖοι, Dečev TSR p.293) from which it might have been created the hydronym *Μένδαπα> Μάνδεπα. The only sure thing here is that the root Mand-/Mend- or Mant-/Ment- occurs mostly in Rhodope and in eastern Macedonia (Dečev TSR s.v. Μανδαραί in Macedonia, Μαντα+var. by Hypate, Larissa, Philippi, Prosočani, Μέντης + var. in Ciconia a.s.o.). In Ortel.Thrak, in the list of unlocalized names, we may find a new variant of this toponym (in Latin transcription): Mandepsa, which could change our etymological orientation. Vaguely localizable in Rhodope.
145.27. Θαρσάνδαλα. Hapax. From Beševliev's etymological propositions (PKN p.127) I retain the relation between the ending -δαλα and the first formant of the name Δαλάταρβα (see infra), of celtic origin in his opinion . But the root Δαλα-, Δαλη- is also well attested in the Thracian space (see Dečev TSR p.114: Δαλαζελμις, Δαλαιτραλις, Δαλαΐλος etc.) where it could come from ie.*dhe-"to settle", also found in -dava and -dama (see Uscudama and Sadame Dečev TSR s.v.). It could consequently be a settlement in the territory of the Darsioi (Dečev TSR p.120: Δάρσιοι, Δερσαῖοι), whose name, which ment – most probably - "the Courageous ones, the Dearers"(< ie.*dhṛsu, see sl.drŭzŭ, engl. to dare etc.), was genetically related and very close to gr. θάρσος (dialectal variant of θρασύς). Θαρσ- could then be either a Hellenization of the local name or a parallel variant of Dars(i)oi in a dilaect which knew the consonantic mutation. Naturally split into Θαρσαν- and -δαλα, I am very confident that the name, having its in the Thracian plural of the genitive, must be translated as "the settlement of the Darsioi/Tharsai".
145.28. Δένιζος.Hapax. No indication for localization. Built with the suf.-ζ- (-ζος,-ζα), frequent in the Thracian toponymy, it might contain the panthracian Dine(n)s, Dinis (see Dečev TSR p.137-138), but I have serious doubts about that.
145.29. *Τόπερον (τόπαρον V, τὸ πόρων V). Mentioned also in the descriptive text [144,16-19: ἔστι δέ τις ἐν ῾Ροδόπῃ πόλις ἅρχαία Τόπερος ὄνομα] the town is well known from other sources too: Strabon, Pliny the Elder, TP, GR. It was in Rhodope, today ruins on the eastern bank of Mesta, near is mouth. Discussion and etymologies by Beševl. PKN p.128.
145.30. Δαλάταρβα (δαλάσαρβα A). Because it also appears in the list of Haemimontus (147,39), the toponym could be on the border between the latter and Rhodope. For its first part, Δαλα-, see supra Θαρσάνδαλα. The second formant, -ταρβα, could also be Thracian (see Ταρβος/Τερβος and tarabostes, Dečev TSR p.490 and 489).
145.31. Βρέ. Beševl. PKN p.128: could be the same as Βρέα - by Steph.Byz 185,7 and in some inscriptions (see Dečev TSR p.86) - an Athenian colony on the Aegeic shore of Thracia, therefore in Rhodope.
145.32. †Κουσκάβιρι (κουκάσβιρι A) and
145.33. †Κούσκουλις (κουσκούλης V). Here at the end of the list of Rhodope, both appear in the same order at the end of the Moesian list (under the forms Κούσκαυρι and Κούσκουλι). Beševl. PKN p.71-72 and 128 considers them as toponyms, and the Latin resonance of the second makes him propose an etymon Quisquiliae (p.128), phonetically unacceptable since lat. qui- is constantly rendered by Procope by κυ(ι)- or by κουι-(see Κυντῶν = Quintorum, Κυίντου = Quinti etc.). They look to me, mostly because of their position but also due to their aspect, a marginal note in a Latin, afterwards included in the text, something like: conscabri (or conscrivi) cum scoliis "full of [or I also transcribed the] scholia (or scratches)", or anything else of this kind (cf. cuscolium).
Under the Title Θρᾴκης(146,1)
146.2. Βόσπαρα. Hapax. Probably on the road Serdica-Philippopolis, above Bessapara (see the next toponym). The name comes from the IE root *gwōu-"cattle, ox" and can be either integrally Thracian, or a Greek-Thracian hybrid, - see βου-δάθλα and [in the dative plural?] Βουσί-παρα, a village near Pizos in the same region: Dečev TSR p.545 and 83; see also gr. Βόσπορος. It corresponds topographically to Vakarel, a small town and peak, east from Sofia, whose name is inherited from rom. văcărel "the little cowboy", maybe a perpetuation of the ancient meaning.
146.3. *Βεσάπαρον (sic Beševliev, codd. βεσούπαρον V βεσύπαρον A) PKN p.128: the same as Bes(s)apara "The City of the Bessians" from IA (Bessapara) or IGB 947 (Βεσσαπαρηνοι). Today Sinitovo (by Dečev TSR p.55: Bešikara, a name not mentioned by Delir. IGT and which I couldn't find anywhere else).
146.4. Καπιστούρια( καπιτούρια A). Hapax. The first component occurs also in Capidava and Capora (Dečev TSR p.226); cf. also Κοπούστορος that Dečev (TSR p.253) takes for a variant of this name. The second component, -στουρια, is (following Beševl. PKN p.129) the same as -styrum in Getistyrum "locus possessorum" (v. et infra 147.33. Γητριστάους).
146.5. Βηρίπαρα ( βερίπαρα A). Beševl. PKN p.129 prefers, with no motivation, the form given by the manuscript A and supposes that the town lied somewhere between Thracia and Moesia, because its name also occurs in the Moesian list (148,1). But the referred sequence of names - between Κιστίδιζος (148,12) and ᾽Ασίλβα (148,24) - is, as I shall soon sustain3, an interpolation from the list of Haemimontus. Therefore I think that there were two homonymous places: one in Thracia, the other in Haemimontus (cf. other such homonyms like Τομις, Βριπαρο(ν) a.o.)
146.6. Ισγίπερα. Hapax. Beševl. PKN p.129 considers it of celtic origin (from isc "water"). Taking into account that it is built with -περα (used as a variant of -παρα in other Thracian names too: Τόπερα, Πρισκούπερα etc.) we must agree that it is either a Celtic-Thracian hybrid or only Thracian (see the same unusual group -σγ- in Ἄσγαρζος and Ἀσγίζους). It was, almost surely, on the central road of Thracia, along the Hebrus river, which makes its celtic origin less probable.
146.7.*Ζορμη (ὀζόρμη V, ὀζάρβα A). Beševl. PKN p.129 identifies the place (without correcting its name) with Zyrmis (TP 8,1) and Zirmis (GR 191,6) – to which let me add also Syrnota from IH 568,5 -, on the river Syrmus (Plin NH 4,50), today Strjama, a northern tributary of the Maritsa(Hebrus), therefore localizable still on the central road of Thracia.
146.8. [†Βηρηΐαρος] (βοριίαρος A). It is a marginal variant of Βηρίπαρα, whose original form was probably *Βηρίπαρος; the author of this note made it in order to record a masculine variant in -ος of Βηρίπαρα, found by him in another manuscript. After being altered and nomore recognized as a variant (the group ΗΙ was miswrote ΙΠ in uncial letters), it has finally been included in the text. Beševl. PKN p.129 takes it for a stand-alone name, but I don't agree. The archetype was surely more severely damaged in this area of the page, as proven by the following two names.
146.9. *Παρεμβολί (codd. ταμονβαρί). IH mentions in this range, I mean between Syrnota and Carassura, two names: Paremvole and Cillio. The first of them is originally the gr. παρεμβολή "military camp", pronounced at the time paremvolí. The number of syllables, the position of the accent, the presence of the group -mv-/-nv-, the vowel context, the confusion Π-/Τ- in the Procopian text (see e.g. 121.23 Τρισκίανα for Πρισκίανα) but most of all its collocation persuaded me that the original must have had here, probably from a Latin source, the form *Paremvoli, corrupted afterwards in Ταμονβαρί. Other etymological references (see Beševl. PKN p.129) are much less probable.
146.10.*(εἰ)ς Κελ(λ)ήνας (σκελήνας A, σκέμνας V). All modern authors, beginning with Haury himself, preferred the variant σκέμνας found in the ms. V, based, of course, on no positive argument, except for that that V seems to be the better manuscript. In my opinion it is quite the other variant which has the best chances to be original. We must first observe that the plural accusative is a strong indication that the initial σ- stands here for the preposition εἰς, as in Ζδέβρην (see below, at 147.25). The rest of the name, *Κελ(λ)ήνας, is nothing else but an ethnicon from Cellae, a place which lied precisely in this spot (see also the previous name), namely Orizovo, by Čirpan (former Kara Orman-for Dečev's readers): it is the same as Cillio from IH and from other sources as well, and its name ment "sources, wells", from ie. *gwel-no- (see Dečev TSR p.238 s.v Κέλλαι) or *qwel-no-. Among its attestations we may also find derivatives built with the –enos suffix for ethnicons: Κελληνος (Dečev TSR p.238 s.v). We can even document that in this word the double -λλ- sometimes spelled with a single l: in the compound Σαλδο-κεληνος (a transliteration, together with Salte-caputenus, of the much more frequent local name Σαλδο-βυσσηνος) by Dečev TSR p.412-413. The use of the derivated adjectives instead of the bear names became a custom by late Latin authors: see passim Zapparena civitas for Zapara, Naissitana civitas for Naissus, Odyssitana for Odyssos etc.). Other explanations, less probable, (starting from var. Σκέμνας) can be found in Beševl. PKN p.
146.11. *Καράσουρα Sic Beševl. (PKN p.130), codd: καράσθυρα V, καράσυρα A. Known also from other sources (see Beševl. ibidem). Today ruins on the hill Kale, between the villages Rupkite and Svoboda by Čirpan.
146.12. *Πίζος Sic Beševl. (PKN p.130), cod: πίνζος V. The wrong form found in V is – in my opinion - a misinterpretation of an original πῖζος, because at least one of the copysts used to have big problems in understanding accent signs (see e.g. Τιγᾶς instead of Τίγρας or ᾽Αντωῖνον for *᾽Αντωνῖνον 146.43.). Known from other sources as well (see Beševl. ibidem). Today, ruins by Dimitrievo (former Čakărlar).
146.13. Τουλεοῦς (?). Etymological suggestions by Beševl. PKN p.130. It was near Arzus, where IB mentions a mutatio Palae. It doesn't seem impossible to me to deal here with this very form indeed. From Palae (in a Latin original?) *Τουλε(+οῦς<?) presents two mutations, both very common in De Aedificiis: Π > Τ (see Τρισκίανα for *Πρισκίανα) and α>ου or viceversa (see Βεσούπαρον for *Βεσάπαρον or Τούρριβας instead *Τούρριβους).
146.14. ῎Αρζον (ἄρξεν A). Well attested toponym whose name, corresponding to gr. ἅργός, ment "(shiny) white" (see Dečev TSR p.24 and Beševl. PKN p.130). Its collocation was on the homonymous river, identifyed by Dečev and Beševliev with Sazli-dere (from tk. saz "reed"). But some other quite recent maps name the river Sütlü-dere (<tc.sütlü "milky"). Today ruins by Kalugerovo.
146.15. *Κάστρα Ζάρβα (sic Beševl. PKN p.131). Codd: καστράζαρβα V, καστρόζαρβα A. The same as Castra Iarba in IA 231,5 and mansio Castozobra in IB 568,11, see Dečev TSR p.176 and Beševl. loc.cit. Today ruins not far from Orjahovo (former Saranli).
146.16. †Ζωσίτερσον (ζωσίτερπον A). Beševl. PKN p.131 confers mutatio Zesutera finis Europae et Rhodopae from IH 602,1-2, adding that the two toponyms cannot be topographically identical. In this spot, namely between Arzos and Burdipta on the main road, the last place in Thracia mentioned by IH 137,1 is Subzupara (probably pronounced *Suzupara,-um) which is quite close to our name. The ending -ρσον (-ρπον) is suspect.
146.17. Βέργισον (βέρτισον Α). Hapax. The root occurs also in Βεργέπολις, town in southern Thracia belonging to Abdera, after St.Byz. 163.13. It was probably in the Rhodope mountains, on the border between Thracia and Rhodope.
146.18. Δίγγιον V or Δύνγιον A. If the better variant is that in the ms. A, then we could think to a corrupt form of Δύμη, Dymae (Dečev TSR p.160), town on the border between Thracia and Rhodope, or a late diminutive of this one, *Δύμιον. Other explanations by Beševl. PKN p.131.
146.19. Σάκισσος ( σάκισος A). Hapax. The root Sac(i)- occurs also in both the Daco-Moesian area(Saci-dava), and the Macedo-Thracian one (for ex. Σάκος in Pieria). No indications for localization.
146.20-24. The sequence κουρτουξοῦρα-ποταμοῦκάστελλαν εἰσδίκαια τὸ ἐμπόριον ταύρων κεφάλων V, respectively κουργάξοῦρα ποταμοῦ καστέλλαν εἰσδίκαια τὸ ἐμπορεῖον ταύρων κεφάλων A, seems to me to be more like a short Greek text than a string of names. Its actual form in Procope's modern editions (Κουρτουξοῦρα-Ποταμοῦκάστελλον -Εἰσδίκαια - τὸ ἐμπόριον Ταυροκεφάλων) is a creation of Maltretus, kept by Haury unchanged.
146.20-21. Κουρτουξοῦρα, more probably *Κουρτουζοῦρα (taking into account the common confusion Ξ/Ζ in Greek manuscripts) than *Κουρτουσοῦρα, as restores Beševl. PKN p.131, is surely a hydronym (see Dečev TSR 470 s.v.-σουρα), and this is why its association with 21. ποταμοῦ comes by itself. The full name, in the genitive, would then be *Κουρτουζοῦρα ποταμοῦ, from a nominative *Κουρτουζοῦρας ποταμός.
146.22-23. κάστελλαν (καστέλλαν A) a word obviously corrupted, even if the final -ν appears in both manuscripts. In order to correct it Haury chose to modify the previous -α- to -ο- (proposing κάστελλον), but I think this form has another better explanation:
In a manuscript which had the names already ranged in columns, this final -ν was writen between the lines and much bigger, as we often find in the Greek cursive:
κάστελλα |
|
εἰσδίκαια |
Yet it belonged not to the first row, where κάστελλα is a correct plural, but to the second one which becomes this way εἰσδίκαιαν, a form to be naturally split in εἰς Δίκαιαν "towards Dicaea".
146.23.-24. The next syntagm, τὸ ἐμπόριον ταύρων κεφάλων, fits very well as an apposition to the place name shown above, Dicaea. The city of Δίκαια, known from Herodot, Strabon, Plinius etc., was not far from Abdera, right near the promontories which border the entrance in the Βιστονίς λίμνη (the modern Vistonis limni or Portolago bay), fjords-like on the map, named by Strabon αἱ Θασίων κεφαλαί. I found no other better association for ταύρων κεφαλαί mentioned here by Procope.
Therefore our whole fragment must have had the following original form: *Κουρτουζοῦρα ποταμοῦ κάστελλα εἰς Δίκαιαν, τὸ ἐμπόριον ταύρων κεφάλων and mentioned, without naming them, some castles along a river *Kurtuzuras which flowed (from Rhodope) towards Dicaea. Other explanations by Beševl. PKN p.131-132.
146.25. Βηλαϊδίπαρα (βολαοδύπαρα A). Hapax. Beševl. PKN p.132 proposes a hybrid (ad) Villa(m) Idipara, unprobable on phonetical considerents: latin -ll- is very solid and it is very constatly transcribed by Procope by -λλ- (cf. e.g. Καστέλλιον, Λάβελλος Κελλιριανά,Βαλλεσιανά etc.). This very root, with -λλ-,Βήλλα exists too, at120.21, and there it can be indeed explained that way. But in this case I think we deal with something else: apparently there was in some Thracian names a lexical component or suffix -id(i)- which used to occur at the end of the first part in compound names. Something very similar we find in an epithet of Heros at Nicopolis ad Istrum (Dečev TSR p.43 - 44): Βασκιδιθιας, analizable in Βασκ(ι)-ιδ(ι?) + ιθιας or Βορα-ΐδης. This component could even be a correspondent of the Greek suffix -(e)id- from the IE root *wid-/*weid- "to see, to look (like)". So, our name must be Thracian, of the form: adjective in ιδ(ι?) + παρα.
146.26. Σκίτακες. Hapax. Dečev TSR p.459 compares it with Σκίθαι, a town near Potidaea, in Pallene, and with celt Scitiacum (Holder 2,1398). Not localizable.
146.27. Βέπαρα. Hapax. The first formant occurs also in Be(o)-dizos (IH 601,9 and 570,1) in Europe (between Resisto and Apris, today Bunarli, apud Dečev TSR p.46).
146.28. Πουσινόν ( Παυσηνόν A). Hapax. Beševl. PKN p.132: can be Thracian or celtic. Not localizable.
146.29. ῾Υμαυπάρουβρι and 146.30. Σκαριωτασαλούκρα are discussed by Beševl. in PKN p.13 and 132, where he admits that both are corrupted. The first name is to be split in ῾Υμαυπάρου (= ῾Υμαυπαρα with the confusion ου/α) and –βρι. The Bulgarian scholar compares it with the name Siamaus, but cf. also Impara (by Porsulae). The ending -βρι could represent either the thrac. βρία "city" (cf. Βρέ above) or, together with the beginning of the following name, *Βρίσκα (Beševl. PKN p.13), to be bound to the late variants Brizice (IA 331,8), Berozicha (IH 602,11), Brenzici (TP 8,8) and Prindize (Guido 535,16) of the name Brendice (IA 322,1), a town near Porsulae, if not even another name of this very town, today Kapdžilar (see Dečev TSR p.86). What follows can be anything: from *(ad) Rotas alacra(s), as proposed by Beševl. in PKN p.13, until *Οὐλπ. Τρα. [Αὐγούστας] (see next name) , *εἰς καρυώτα σαλουβρα or even *᾽Ισκαριωτα(ς) as an expression of some scribe's trouble.
146.31. Αὐγούστας (see Beševl. PKN p.132). Identified by Litzica 71 with Augusta Traiana and by Beševliev with Hisarja Baths, in both cases in northern Thracia.
146.32. Οὐρδαούς can be, after Beševl. PKN p.132-133, an alteration of the name Orudisza (ad Burgum IA 175,4) which is in Haemimontus, at Goljam Manastir, by Topolovgrad (former Kavakli).
146.33. Τοῦ ἁγίου Τραϊανοῦ (scil. castellum). Beševl. PKN, after Delhaye Origines 241-242, says that there is a therapeute saint Traianus, worshipped mainly in Macedonia, but this cannot offer any indication for localization.
146.34. Δέρταλλος Beševl. PKN p.133 compares it with the celtic Dertona, Dertosa. Cf. still Τέρτα by Ptol. 3.10.7, near Serdica, on the road to Philippopolis, and Μαμού-δερτα, a town in Bithynia (Dečev TSR p.284), both Thracian. It is not at all impossible that the original form would be *Δέρτα ἄλλος, as it happens with Κατταφέτερος in Dardania (120.10), which must be read certainly *Κάτταρ(ος) ἕτερος, and in this case the basic name would be *Δέρτα, the same as Τέρτα (v. supra) with late consonantic mutation as in Δορδᾶς by Procope's time, but *Δορταζα earlier (see below, at 147, 12-19). Not localizable.
146.35. Σολβανοῦ (scil.castellum) Beševl. PKN p.133: from a silvanus or Salvenus. Not localizable.
146.36. Βάσκον Beševl. PKN p.133 supports its celtic origin. But there is no need for that, since it occurs in Thracian too: see the anthroponym Bascila(s)+var., quite frequent in Edonia (around Drama and Philippi). There is also a Βασκιδιθιας, epithet of Heros attested at Nicopolis ad Istrum (Dečev TSR p.43 - 44).
146.37. Ζίγκυρο. Beševl. PKN p.133: possibly Thracian. I doubt. There is no match in Thracian onomastics. The final part would render a Latin -quirus, but I found nothing to fit the beginning. Not known and not localizable.
Under the Title Αἱμιμόντου(146,38)
146.39. Ζημάρκου Beševl. PKN p.133: = Zemarci (scil. castellum or fossatum), or maybe praedium. There are more than one Zemarcus (Zimarcus) in this period , all of them Thracian (see Russu Elem. 179). Not localizable.
146.40. Κηριπάρων (sic l, /ηρι/πάρον V). Beševliev (Beševl. PKN p.133) supposes it is a Gen.pl. and compares it with ἐν Κειρπάροις (IGB 2338, a town in the valley of the Strymon river), with Κηριβωστ[η]νοι (IGB 270, a village in northern Haemimontus) and with Κεῖρις, the name of a cave recorded by Dio Cassius 51,26,3 in Scythia Minor. Phonetically, this name is identical with Κειρπαροι, which survived until the modern era (Ciropol), even if they don't seem to concord topographically. Therefore I think that, like Βηρίπαρα above and like other toponyms (e.g. Βρίπαρο(ν) by Procope, or Buridava, Sucidava etc. in other authors) we deal here with two different places having the same name. In order to localize it we might better consider an orientative identity with *Κηριβωστα (in Κηριβωστηνοι) in northern Haemimontus.
146.41. Κασιβόνων Beševl. PKN p.133 = gen.pl. and compares it with the name of the celtic god Casebonus, found in a Latin inscription from Lomnica (Trăn). We could indeed have here a late celtic community which preserved their old god's name in a toponym. And yet see also Βασίβουνον, because it can be as well a graphic confusion β/κ, rather common in Greek manuscripts.
146.42. Τὸ Οὔκου. This name may be compared with the hydronym Ucasus (river in the Succi pass), with the anthroponym Ucus Dydigis from Salonic (Dečev TSR p.348), both of them pleeding for its Thracian origin.
146.43.*᾽Αντωνῖνον Sic Beševl. PKN p.134, cod. ἅντωῖνον, following Philippide Orig.435,468. It would be adjectival, in a syntagm like ad fundum ori saltum Antoninum. Not localizable.
146.44. *Γεσιλᾶ φοσσᾶτον (sic Beševl. PKN p.134, cod. γεσιλαφοσσᾶτον), Gesilae fossatum in the -ᾶ genitive of the foreign words. The name is surely gothic. Not localizable.
147.1. Χεροῖνον After Beševl. PKN p.134, very similar to the Greek dual χεροῖν of the noun χεῖρ,-ός "hand" which makes him see here a possible tavern name, meaning "To Both Hands", which is very doubtful. Better see a similar context three lemmas above, in ᾽Αντωῖνον instead of ᾽Αντωνῖνον. Supposing that we deal here with the same mistake, we might reconstruct a *Χερονῖνον (= Cheroninum < Cheronius?). Not localizable.
147.2. Προβίνου (scil. castellum): Beševl. PKN p.134. See next name. The toponym also occurs at 147.35, also in Haemimont. This is why Beševliev supposes, with good reasons, that it perhaps belonged to a place at a crossroads.
147.3. Τοῦ ἁγίου Θεοδώρου (scil. castellum). Beševl. PKN p.134 quotes Delehaye, Orig. 241 who says that Sct. Theodorus from Euchaitia and Sct. Iulianus from Cilicia were already well-known in early antiquity. The name also occurs further, at 147.26. I think that both Προβίνου and Τοῦ ἁγίου Θεοδώρου were on an important road (very probable on the main road of Thrace), close one to the other, and that each was also the ending point of lateral roads that Procope follows as well. So, he mentions those two places first as stations on the main road and second as ending points of the other two lateral but important roads.
147.4. Βουρδέπτω Beševl. PKN p.134-135: well known from other authors too. Today the hill Hisar by Svilengrad (former Mustafa Paša). The first part of the name comes, no doubt, from the IE *bhordh- "ford" (germ. Furt, sl. brodu, gr.πορθ-μός) and indeed, today as maybe in ancient times, a branch of the main road crosses the Hebrus to follow its right bank on to the see.
147.5. Ρακούλη. Compared by Beševl. PKN p.135 with ῾Ρακώλη - the name of the land owned illo tempore by the pygmies, mentioned by St.Byz 368,11 and placed by Pliny NH 4,44 north from Odessos - but also with ῾Ρανούλη, found in a protobulgarian inscription, the same as Ranilum in TP 8.2 şi GR 4,7.
147.6. Τοῦ ἁγίου ᾽Ιουλιανοῦ Not known and not localizable. For the name, cf. Τοῦ ἁγίου Θεοδώρου above.
147.7. Τζιταετοῦς Beševl. PKN p.135, after Tomaschek and Philippide, thinks it comes from a lat. Civitas Vetus or Cittá Vetus. Though tempting indeed, this etymology seems to me doubtful, both for phonetic reasons (the v in the hypothetical Vetus wouldn't fall in initial strong position) and because we are here in the middle of the hellenophone territory, where Latin names can hardly be expected. Much more chances to be true has the explanation (εἰς) Τζίττα ἅετούς "To Tzitta's eagles (=aquilae)", a name very well suited for a tavern, built from that of the contemporary general (see Russu Elem 106 sqq. and footnotes). Could it be in relation with the name of the later and modern town Aitos?
147.8. Βηλαστύρας Beševl. PKN p.135= pl.acc.? Dečev TSR p. and the others translate "white place", while Beševl. PKN p. thinks of the transliteration of a Latin Villa(m), both without real chances. Due to the different functions assumed by vowels ε and η in Thracian onomastics, there are few cases where they really alternate in variants of the same name, quite like in English e and ee. Cf. Βελεδίνα (infra) and the next name.
147.9. Γετρίνας. It is not to be bound to Γητριστάους as does Beševl. (PKN p.135). As I said above, the alternance ε/η is forced. It looks to me that we deal here with a lat. Petrinas, with the usual graphical confusion Γ-/Π- (see also Γριγκιὰπανα instead of *Πρινκιπιὰνα at 123.41 in Dacia Mediterranea)
147.10. Βρέδας cf.Βρεντ-?
147.11. Βῆρος
147.12-19. The string of names Θωκύωδις, Βία, ᾽Αναγογκλί (ἅναγονκλί V), Σούρας, Αὐθιπάρου, Δορδᾶς, Σαρμαθῶν, Κλεισοῦρα lies today in a column which belonged, in the original, to the title Θρᾴκης, as I shall soon show3 in an article. Taken as they are, all we can extract from these names is that they are to be placed somewhere at the mountain (see Κλεισοῦρα). It is also obvious that ᾽Αναγογκλί is rather a corrupted form. But, if we join these names together, in string of scriptio continua, and then we split them again, with no letter changed(!), we might obtain the following much acceptable toponyms:
*Θωκύωδις βία is, I think, a Greek transcription of a Latin (ad) Thocyodis via(m), in the genitive of the type custōs,- ōdis (observe the length respected!), from a Thōcyōs which, at its turn, may render a dm. *čoku-(cf. rom. ciucă "peak"). It may as well be the same as Θουκυ- in Θουκυσιδαντική, the name of a strategy in Thrace - a possibility that worths further investigations. As for the way the local people used to built road-names, see Τραϊανοῦ τρίβος, placed by Theoph. Simoc. 8, 4, 3-4 in this very area.
*῎Αναγον (sc.κάστελλον) could be a neuter participle from ἅνάγω, meaning consequently eminens (castellum)
*Κλισοῦρα Σαυθιπάρου. Σαυθίπαρος is the autochthonous, bessian, form of the Hellenized name Σευθόπολις, a very well known town in northern Thracia. For Saut-=Σευθ- see Dečev TSR 421 who quotes a Dines Sautis of Bessic origin (CIL 10, 3590 at Misenum), to whom I can also add Bitus Sautes (from CIL 3,8095). Though the town existed nomore in Procope's times, its name could have very well been preserved in that of the pass in the proximity. Both town and pass were in northern Thrace, near today Koprinka, at south of the Šipka pass, where there have been found the ruins of Seuthopolis, today submerged under the waters of an accumulation lake of the local hydro-electric plant, after being completely studied by the Bulgarian archeologists. The following name happily confirms this hypothesis.
Δορδᾶς cannot be related but to *Δορταζα, a place name reconstructed from Δορταζηνος. This theonym is attested at Krăn, only 5 km. north from Seuthopolis, on the road towards Şipka pass (see Dečev TSR 151).
*Σαρματῶν (Sic Beševl. PKN p., cod.σαρμαθῶν) Κλεισοῦρα could be therefore the Šipka pass itself, especially if we recall that Pliny (NH 4,41) mentions in this very spot, right at the exit from the pass, a Sarmatian enclave that Iordanes indirectly confirms, when he tells us (in his Getica 101) that Traian built Victoriae Civitas to celebrate the memory of his victory over the Sarmatians. Or, this town of Traian is Nicopolis ad Istrum, few kilometers north from Šipka pass.
You may find however other explanations of these names at Beševl. PKN p.135-136.
147.20. ῾Υλασιάναι with a curious nom.pl. The names in -ιανα are derivated, as Beševliev shows, in PKN p.48-50, from anthroponyms (Αὐριλίανα from Aurelius, Βαλεντινιάνα from Valentius, Εὐτυχιανά from Eutychis, Λαμπωνιανά from Lamponius etc., a very well represented group).
(This translation will be continued soon. For the rest of the text you may consult its full Romanian version)
Notes
1. In Georgius Cyprius 62, 1263; H. Gelzer, Ungedrückte und ungenügend veröffentlichte Texte der Notitiae Episcopatuum, Berlin,1900, p. 552, 158-160 and G. Parthey, Hieroclis Syecdemus et Notitiae Episcopatuum, Berolini, 1866, 103, 52-54; 107, 148; 183, 54-55; 245, 47-48.
2. Veselin Beševliev, Die Protobulgarischen Inschriften, Berlin, 1963
3. In an article not yet published which I shall soon submit to your criticisms, I support the hypothesis that one of the biggest mistake made by the copysts was the misplacement of the titles, which led to assigning names to wrong provinces. The pattern of these mistakes is recognoscible, so that the original order can be hypothetically restored. Some of references here send to this future article.