I n 1864 Flóris Rómer, the pioneer of the Hun- garian archaeology and epigraphy when he published the Aesculapius-altar from Annama- tia (Baracs-CatNr. 120) wrote that he hardly knew Greek inscriptions in Pannonia with the exception of a bilingual sarcophagus from Brigetio (CatNr.74 = CIL III, 4327) and the vas diatretum from Szekszárd (CatNr. 121) 1 . Meanwhile the number of the Greek inscriptions found in the province and the neighbouring Barbaricum signif- icantly increased therefore it was necessary to pub- lish all of them together at the same place 2 . In my paper I wish to examine the results of this work and the relationship between the Greek language and Pannonia. First of all I must emphasize that sometimes it is very difficult to distinguish the fragmentary Greek and Latin inscriptions, esp. in the case of the graffiti 3 . For example, I wish to present a late Roman fragmentary stone inscription (RIU 122) (most likely a grave stone) from Savaria (Fig. 1) 4 . Only the last letters of the six lines of the framed marble plate remained: - - - - - - / [ - - -]I / [- - - ]S / [- - - ]N / [- - -]AES / [- - -] PEL / [- - -] OY / - - - - - - The inscription can obviously be dated to the 4th century and it most likely belongs to the numerous Christian sepulchral inscriptions of Savaria (RIU 75-86, S 26) 5 . Another fact is that the inscription is written in Latin characters. The only questionable line is the last one where the letters OU can be seen. The diphtong can belong to a Latin and a Greek sentence-expression as well. The letter Y was used in the Pannonian inscriptions almost only in the case of Greek (Oriental) per- sonal and place-names. The Greek dipthong OY was almost always transcripted as U in the Latin. In Pannonia there is no similar evidence for this use but it cannot be excluded that this diphtong can occur in Latin inscriptions as well. Among the Graecanica phenomena of the Latin inscriptions in the CIL III there are some cases where Y was used instead of the letter U (CIL III p. 2573, 2577, 2677). To this group belong those inscriptions as well where in the Latin text the Greek diphtongs AY, EY, OY were written as in the Greek (Ayr(elius): CIL III, 8935, 6594, Heytycinus: 4318, OY pro U: CIL III, 218) 6 . This phenomenon can be observed always in the case of Latin inscriptions found in the East or Greek (Oriental) persons in the West. The other, more probable possibility is that this is a bilingual titulus where in the last line a Greek term was written. In Pannonia there are sev- eral bilingual inscription but this would be the 2nd Christian one. Beside the numerous Chris- togrammata 7 the only exception is the grave stone from Rákospalota (CIL III, 13382=Kovács, o.c., Nr. * Pármány Péter Catholic University, Hungary. 1. Diary of Flóris Rómer. Manuscript. XIII, 1864, 104 (with drawing), Archaeologiai Közlemények 4, 1864, 47-48 = CatNr. 120. 2. KOVÁCS, P., Corpus Inscriptionum Graecarum Pannonicarum. (Hungarian Polis Studies 8), Debrecen 2001. 3. Cf. KOVÁCS, o.c., 5. 4. BALLA, L.; P. BUOCZ, T.; KÁDÁR, Z.; MÓCSY, A.; SZENTLÉLEKY, T., Die römischen Steindenkmäler von Savaria, Budapest 1971, 123, Nr. 203. 5. GÁSPÁR, D., Christianity in Roman Pannonia. An evaulation of Early Christian finds and sites from Hungary (BAR IS 1010), Oxford 2002, 122-125, Nr. 51.I.a-f, II.b-g. 6. MIHAESCU, H., La langue latine dans le sud-est de l’Europe, Bucarest, Paris 1978, 183-184, Nr. 130. 7. FEHÉR, B., Lexicon epigraphicum Pannonicum, Budapest 1997, 228. GREEK INSCRIPTIONS IN PANNONIA PÉTER KOVÁCS* ACTA XII CONGRESSVS EPIGRAPHIAE GRAECAE ET LATINAE, pp. 785-792