Opera Instituti Archaeologici Sloveniae 23 Poznoantična utrjena naselbina ' .. . Tonovcov grad pri Kobaridu Late Antique fortified settlement Tonovcov grad near Kobarid V.; . Naselbinski ostanki in interpretacija Settlement remains and interpretation Slavko Ciglenečki Zvezdana Modrijan Tina Milavec Zbirka / Series Uredniki zbirke / Editors of the series opera instituti archaeologici sloveniae 23 Jana Horvat, Andrej Pleterski, Anton Velušček Slavko Ciglenečki Zvezdana Modrijan Tina Milavec Poznoantična utrjena naselbina Tonovcov grad pri Kobaridu. Naselbinski ostanki in interpretacija Late Antique fortified settlement Tonovcov grad near Kobarid. Settlement remains and interpretation Recenzenta / Reviewed by Urednica / Editor Prevod / Translation Jezikovni pregled / Language Editors Tehnična ureditev / Technical Editor Oblikovanje ovitka / Front cover design Fotografije / Photographs Računalniški prelom / DTP Priprava slikovnega gradiva / Preparation of illustrations Izdala in založila / Published by Zanju / Represented by Glavni urednik / Editor-in-Chief Tisk / Printed by Jana Horvat, Mitja Guštin Zvezdana Modrijan Sunčan Patrick Stone, Tina Milavec Sonja Likar, Alan McConnell-Duff Mateja Belak Tamara Korošec Slavko Ciglenečki Mateja Belak Drago Valoh, Lucija Lavrenčič, Mateja Belak, Tamara Korošec Inštitut za arheologijo ZRC SAZU, Založba ZRC Oto Luthar, Jana Horvat Aleš Pogačnik Collegium Graphicum d. o. o., Ljubljana Izid knjige sta podprla / Published with the support of Javna agencija za knjigo RS, Znanstvenoraziskovalni center SAZU Fotografija na ovitku / Front Cover photo Slavko Ciglenečki CIP - Kataložni zapis o publikaciji Narodna in univerzitetna knjižnica, Ljubljana 904(497.4Tonovcov grad)"652" CIGLENEČKI, Slavko, 1949- Poznoantična utrjena naselbina Tonovcov grad pri Kobaridu : naselbinski ostanki in interpretacija = Late antique fortified settlement Tonovcov grad near Kobarid : settlement remains and interpretation / Slavko Ciglenečki, Zvezdana Modrijan, Tina Milavec; sodelavci Benjamin Štular, Saša Čaval in Ivan Šprajc = with contributions of Benjamin Štular, Saša Čaval and Ivan Šprajc; [prevod Sunčan Patrick Stone, Tina Milavec ; fotografije Slavko Ciglenečki]. - Ljubljana : Inštitut za arheologijo ZRC SAZU, Založba ZRC, 2011. - (Opera Instituti Archaeologici Sloveniae ; 23, ISSN 1408-5208) ISBN 978-961-254-331-0 1. Modrijan, Zvezdana 2. Milavec, Tina, 1979258819584 © 2011, ZRC SAZU, Inštitut za arheologijo, Založba ZRC Vse pravice pridržane. Noben del te knjige ne sme biti reproduciran, shranjen ali prepisan v kateri koli obliki oz. na kateri koli način, bodisi elektronsko, mehansko, s fotokopiranjem, snemanjem ali kako drugače, brez predhodnega pisnega dovoljenja lastnikov avtorskih pravic. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the publisher. Slavko Ciglenečki Zvezdana Modrijan Tina Milavec Poznoantična utrjena naselbina Tonovcov grad pri Kobaridu Naselbinski ostanki in interpretacija Sodelavci: Benjamin Štular, Saša Čaval in Ivan Šprajc Late Antique fortified settlement Tonovcov grad near kobarid Settlement remains and interpretation With contributions of Benjamin Štular, Saša Čaval and Ivan Šprajc ZALOŽBA Z R C LJUBLJANA 2011 vsebina Predgovor (Slavko CIGLENEČKI, Zvezdana MODRIJAN in Tina MILAVEC) ...........................................................7 1. Uvod (Slavko CIGLENEČKI)...........................................................................................................................................9 1.1 Geografski oris...........................................................................................................................................................11 1.2 Ime in izročilo............................................................................................................................................................15 1.3 Opis najdišča..............................................................................................................................................................17 1.4 Zgodovina raziskav...................................................................................................................................................25 1.5 Posočje v poznoantičnem času................................................................................................................................33 1.6 Mreža poti (Benjamin ŠTULAR)............................................................................................................................53 2. Terenski izvid (Zvezdana MODRIJAN, Slavko CIGLENEČKI in Tina MILAVEC)...............................................65 2.1 Metodologija dela ...................................................................................................................................................... 67 2.2 Kronologija in faze .................................................................................................................................................... 69 2.3 Stavba 1 z okolico ...................................................................................................................................................... 73 2.4 Stavbi 2 in 3................................................................................................................................................................97 2.5 Sklop cerkva.............................................................................................................................................................111 2.6 Cisterna (vodni zbiralnik)......................................................................................................................................155 3. Stavbni razvoj in kronologija (Zvezdana MODRIJAN in Tina MILAVEC)............................................................163 3.1 Stavba 1 z okolico....................................................................................................................................................165 3.2 Stavbi 2 in 3..............................................................................................................................................................193 3.3 Sklop cerkva.............................................................................................................................................................201 3.4 Cisterna (vodni zbiralnik).......................................................................................................................................211 4. Primerjave (Slavko CIGLENEČKI) .............................................................................................................................215 4.1 Stavba 1.....................................................................................................................................................................217 4.2 Interpretacija cerkvenega sklopa ........................................................................................................................... 225 4.3 Astronomska orientacija cerkva na Tonovcovem gradu (Saša ČAVAL in Ivan ŠPRAJC) ............................. 247 5. Vloga in pomen naselbine Tonovcov grad (Slavko CIGLENEČKI) ........................................................................ 257 5.1 Utrdba Tonovcov grad - pomemben člen poznorimske obrambe Italije.........................................................259 5.2 Tonovcov grad v poznoantični poselitveni sliki vzhodnoalpskega in zahodnobalkanskega prostora.........273 6. Sklep (Slavko CIGLENEČKI, Zvezdana MODRIJAN in Tina MILAVEC)............................................................289 7. Literatura (Zvezdana MODRIJAN)..............................................................................................................................295 contents Foreword (Slavko CIGLENEČKI, Zvezdana MODRIJAN and Tina MILAVEC) .........................................................7 1. Introduction (Slavko CIGLENEČKI) ..............................................................................................................................9 1.1 Geographic description............................................................................................................................................11 1.2 Name and tradition...................................................................................................................................................15 1.3 Site description..........................................................................................................................................................17 1.4 Research history........................................................................................................................................................25 1.5 The Posočje area during Late Antiquity.........................................................................................................................33 1.6 Path network (Benjamin ŠTULAR) .......................................................................................................................53 2. Field report (Zvezdana MODRIJAN, Slavko CIGLENEČKI and Tina MILAVEC) ...............................................65 2.1 Methodology..............................................................................................................................................................67 2.2 Chronology and phases ............................................................................................................................................ 69 2.3 Building 1 and its surroundings .............................................................................................................................. 73 2.4 Buildings 2 and 3 ......................................................................................................................................................97 2.5 The ecclesiastical complex......................................................................................................................................111 2.6 Water cistern (reservoir).........................................................................................................................................155 3. Building development and chronology (Zvezdana MODRIJAN and Tina MILAVEC)........................................163 3.1 Building 1 and its surroundings............................................................................................................................165 3.2 Buildings 2 and 3.....................................................................................................................................................193 3.3 The ecclesiastical complex......................................................................................................................................201 3.4 Water cistern (reservoir) ........................................................................................................................................211 4. Comparisons (Slavko CIGLENEČKI) .........................................................................................................................215 4.1 Building 1 .................................................................................................................................................................217 4.2 Interpretation of the ecclesiastical complex ........................................................................................................225 4.3 Astronomical orientation of churches at Tonovcov grad (Saša ČAVAL and Ivan ŠPRAJC) ........................247 5. The role and importance of the settlement Tonovcev grad (Slavko CIGLENEČKI).............................................257 5.1 The fort at Tonovcov grad - an important part of the Late Roman defence system of Italy.........................259 5.2 Tonovcov grad in the broader Late Antique settlement pattern of the East Alpine and Western Balkans ..............................................................................................................................................................273 6. Conclusion (Slavko CIGLENEČKI, Zvezdana MODRIJAN and Tina MILAVEC) ..............................................289 7. Bibliography (Zvezdana MODRIJAN) ........................................................................................................................295 predgovor foreword Splet okoliščin je pripeljal v zadnjem desetletju prejšnjega stoletja do odkritja pomembne poznoantič-ne naselbine Tonovcov grad pri Kobaridu, v Zgornjem Posočju, v tem prelepem koščku slovenske dežele. Takrat so bile v Sloveniji delno že raziskane prve postojanke iz obdobja prehoda antike v srednji vek, ki so odstrle vpogled v čas obstoja poslednjih bivališč romaniziranih staroselcev. Žal pa jih je bila večina že v preteklosti delno uničena ali slabo raziskana, zato se je zdel Tonovcov grad s svojo kompleksno podobo in izjemno ohranjenostjo pravi izziv za raziskovalno potrditev nakazane, a slabo poznane problematike prehoda antike v srednji vek. Prva sondiranja na najdišču leta 1993 so sprožila zamisel o sistematični raziskavi, saj so tudi kobariški domačini pokazali izredno zanimanje in posluh za naše delo. Raziskava najdišča, ki smo jo vodili sodelavci Inštituta za arheologijo Znanstvenoraziskovalnega centra SAZU iz Ljubljane, se je odlično vključila v sočasne raziskave, ki potekajo tudi v bližnjih območjih Avstrije in Italije: razkrivajo slabo poznano preživetje staroselcev in njihove prve stike z manjšimi skupinami Germanov kot tudi poznejšo slovansko naselitev. Sodobne arheološke raziskave zahtevajo veliko sredstev, saj je znanstveno delo vedno bolj kompleksno in uporablja metode številnih drugih znanstvenih disciplin. Da bi uspešno opravili raziskave in konservacijo nadpovprečno ohranjenih ostankov arhitekture ter jih ustrezno predstavili na terenu, je bila potrebna pomoč številnih institucij in ljudi. Veliko podporo smo imeli v predstojnikih inštituta Janezu Dularju in pozneje Jani Horvat, ki sta omogočila sodelovanje številnih članov inštituta na terenu in velikokrat pomagala pri reševanju nelahkih finančnih težav. Omeniti je treba vsestransko pomoč, ki smo jo dobili od načelnika Upravne enote Tolmin Zdravka Likarja in direktorja Kobariškega muzeja Jožeta Šerbca, ki sta s številnimi spodbudami in pozitivno energijo pripomogla k uspešnemu začetku in razvoju raziskav. Veliko je prispevala Občina Kobarid, predvsem z odkupi parcel in z organizacijo javnih del, za kar gre zahvala županoma Pavlu Gregorčiču in Robertu Kavčiču. Začetne raziskave je omogočilo Ministrstvo za znanost, pomemben pa je bil predvsem prispevek In the last decade of the last century, the turn of events led to the discovery of the important Late Antique settlement of Tonovcov grad near Kobarid in the beautiful Upper Soča Valley. At that time some posts from the period of transition between the Antiquity into the Middle Ages in Slovenia had already been partly researched. They offered an insight into the time of the existence of the last settlements of the autochthonous Romanized people, but unfortunately most of them had either been partly destroyed or poorly investigated in the past. Therefore Tonovcov grad with its complexity and extraordinary state of preservation appeared as a great challenge which could help us solve the indicated but poorly known issues of the transition from the Antiquity to the Middle Ages. The first trial trenches on the site in 1993 led us to consider systematic research, as the locals of Kobarid showed great interest in our work. The excavations, led by the members of the Institute of Archaeology at the Scientific Research Centre of SAZU in Ljubljana, fit very well into the contemporary investigations of the neighbouring areas of Austria and Italy; they uncover the poorly known survival of the autochthonous inhabitants and their first contacts with the groups of Germanic and later Slavic newcomers. Modern archaeological research is very costly as scientific work grows in complexity and uses the methods of various other scientific disciplines. To be able to finish the research and conservation of the extraordinarily well preserved architectural remains and suitably present them on the site, the help of many institutions and people was necessary. We were given great support by the Heads of the Institute: Janez Dular and later Jana Horvat secured us the help of many members of the Institute during fieldwork and often provided assistance in difficult financial situations. We are most grateful to the chief of the Administrative unit of Tolmin, Zdravko Likar and the director of the Kobariški muzej in Kobarid, Jože Šerbec. Their encouragement and positive energy added considerably to the successful beginning and continuation of our research. The municipality of Kobarid contributed greatly by buying off the land and organising public works, for which we wish to thank the mayors Pavel Gregorčič and Robert Kavčič. The initial research was made possible by Ministrstva za kulturo, ki je zagotovilo finančna sredstva in vključilo raziskave v akcijo "kulturni tolar". Posebna zahvala velja tudi tedanjemu direktorju Direktorata za kulturno dediščino Stanetu Mrviču in sekretarju Ministrstva za kulturo Silvestru Gabrščku, ki sta vseskozi podpirala naša prizadevanja in se veselila napredka del. Izkopavanj so se udeležili arheologi Lucija Lavrenčič, Andreja Dolenc Vičič, Primož Pavlin, Janez Dirjec, Anton Velušček, Dragan Božič, Srečko Firšt ter takratni študentje arheologije Veronika Maček, Matija Črešnar, Lucija Šoberl, Dunja Černic, Bojana Rozman, Mateja Ravnik, Miha Mlinar, Rok Klasinc, Barbara Nadbath, Nataša Gomilšček, Bernarda Županek, Samo Hvalec, Samo Sankovič in Januš Jerončič. Tehnično delo so na terenu opravljale Andreja Dolenc Vičič, Lucija Lavrenčič, Dragica Knific Lunder in Tamara Korošec, zadnji dve sta izkopano drobno arheološko gradivo tudi natančno izrisali. Priprave za monografsko objavo najdišča so se začele že med samim izkopavanjem, k sodelovanju pa smo povabili številne sodelavce. Že med delom se je pokazalo, da bo prispevkov preveč za eno knjigo, zato smo opise arhitekture, stratigrafske situacije in izvrednotenje najdišča uvrstili v prvo knjigo (Poznoantična utrjena naselbina Tonovcov grad pri Kobaridu. Naselbinski ostanki in interpretacija), obdelavo drobnih najdb in antropoloških in zooloških ostankov pa v drugo (Poznoantična utrjena naselbina Tonovcov grad pri Kobaridu. Najdbe). Knjigi sta med seboj povezani in tvorita celoto. Besedilo monografij sta pozorno prebrali in predlagali koristne izboljšave Jana Horvat in Andreja Dolenc Vičič. V knjigah so predstavljeni rezultati arheoloških raziskav. Da pa smo lahko zavarovali izkopane objekte, je bila potrebna vrhunsko usposobljena skupina strokovnjakov in delavcev, ki so izvajali konservatorska dela. Najdišče sodi v območje ZVNKD Gorica, zato je odgovorna konservatorka Nada Osmuk, pozorno nadzirala konservatorsko delo in operacionalizirala sklepe konservatorske komisije. Po njeni upokojitvi leta 2007 jo je nadomestila Patricija Bratina. Ekipa Restavratorskega centra Slovenije pod vodstvom Franceta Vardijana (t) in kasneje Jerneja Hudolina je izvajala strokovna dela konservacije in prezentacije arhitekturnih ostankov. Seveda pa vsega omenjenega ne bi bilo brez izvajalcev, ki so jih logistično podpirali in organizacijsko vodili sodelavci muzeja, med katerimi moramo posebej izpostaviti direktorja Jožeta Šerbca, ki je neutrudno pomagal, organiziral in bedel nad potekom del. Številne kovinske najdbe je v Goriškem muzeju odlično konservirala Jana Šubic Prislan. Večino najdb hrani Tolminski muzej, kjer je najdišče že predstavljeno v sklopu stalne zbirke Naplavine obsoške zgodovine. Vsem omenjenim in številnim drugim, ki so po svoje prispevali k raziskavam in kasnejši konservaciji in prezentaciji najdišča, iskrena hvala. the Ministry of Science; an important contribution was also made by the Ministry of Culture. We are also grateful to the then director of the Directorate for Cultural Heritage, Stane Mrvič, and the secretary of the Ministry of Culture, Silvester Gabršček. The excavations included archaeologists Lucija Lavrenčič, Andreja Dolenc Vičič, Primož Pavlin, Janez Dirjec, Anton Velušček, Dragan Božič, Srečko Firšt and (then) archaeology students Veronika Maček, Matija Črešnar, Lucija Šoberl, Dunja Černic, Bojana Rozman, Mateja Ravnik, Miha Mlinar, Rok Klasinc, Barbara Nadbath, Nataša Gomilšček, Bernarda Županek, Samo Hvalec, Samo Sankovič and Januš Jerončič. Technical work in the field was done by Andreja Dolenc Vičič, Lucija Lavrenčič, Dragica Knific Lunder and Tamara Korošec; the latter two also made drawings of the excavated small finds. The preparations for publishing the site as a monograph began already during the excavations. We invited several colleagues to cooperate and very soon it was evident that the number of contributions will exceed the limits of a single volume. Therefore the descriptions of the architecture, stratigraphic situations and the interpretation of the site are included in the first volume (Late Antique fortified settlement Tonovcov grad near Kobarid. Settlement remains and interpretation), while the analyses of the small finds as well as the anthropological and zoological remains form the second one (Late Antique fortified settlement Tonovcov grad near Kobarid. Finds). Both volumes are interconnected and form an integral whole. Jana Horvat and Andreja Dolenc Vičič carefully read the text of the monographs and suggested improvements. In the volumes we are presenting the results of the archaeological research. However, in order to be able to preserve the excavated structures a group of highly trained experts and workers who performed the conservation work was necessary. The site belongs to the area covered by the ZVNKD Gorica, and the responsible conservator Nada Osmuk carefully supervised the conservation works and operationalised the resolutions of the conservation commission. After her retirement in 2007 she was succeeded by Patricija Bratina. The team of the Restoration centre of Slovenia carried out the conservation and presentation of the architecture remains under the supervision of France Vardijan (t) and later Jernej Hudolin. Of course, none of the above would have been possible without the contractors who were logistically supported and organised by the colleagues from the Kobariški muzej, among whom Jože Šerbec must be especially mentioned for his untiring help and the organisation and supervision of the works. Many metal finds were excellently treated and preserved by Jana Šubic Prislan in the Goriški muzej in Nova Gorica. Most finds are kept in the Tolminski muzej in Tolmin, where the site is already presented within the permanent exhibition. To all the above mentioned, and many others who contributed to the research and later conservation and presentation, we are sincerely grateful. 1. uvod 1. introduction 1.1 Geografski oris 1.2 Ime in izročilo 1.3 Opis najdišča 1.4 Zgodovina raziskav 1.5 Posočje v poznoantičnem času 1.6 Mreža poti (Benjamin ŠTuLAR) 1.1 Geographic description 1.2 Name and tradition 1.3 Site description 1.4 Research history 1.5 The Posočje area during Late Antiquity 1.6 Path network (Benjamin ŠTuLAR) 1.1 geografski oris 1.1 geographic description Poznoantična višinska utrjena naselbina Tonovcov grad leži v goratem predelu zahodne Slovenije, ob reki Soči severno od Kobarida (sl. 1.1). Zgornje Posočje, ki sega od izvira reke Soče pod najvišjimi vrhovi Julijskih Alp do Mosta na Soči oziroma Tolmina, spada med najslikovitejše slovenske pokrajine. Soča in njeni pritoki so tu vrezali globoke struge v večinoma karbonatno podlago. Prevladuje dachsteinski The Late Antique fortified hilltop settlement of Tonovcov grad is situated in the mountainous part of the western Slovenia by the river Soča to the north of Kobarid (Fig. 1.1). The landscape of the upper Soča valley, which extends from the spring of the river Soča under the highest peaks of the Alps to Most na Soči or Tolmin, is one of the most picturesque parts of Slovenia. The Soča river and its tributaries have carved deep river Sl. 1.1: Karta obravnavanega prostora z glavnimi poznoantičnimi najdišči. Fig. 1.1: Map of the discussed area with the main late Antique sites. apnenec, iz katerega je zgrajena večina Julijskih Alp. Vmes so pasovi dolomita, rdečega laporja in fliša, na dnu doline pa prevladujejo aluvialni nanosi. Soča do naselja Žaga teče v smeri prečnih dinarskih prelomov, tam pa se obrne v jugozahodno dinarsko smer ter teče ob robu krnskega pokrova. Dolina se močno zoži povsod, kjer reka prečka dinarsko slemenitev, npr. pri Klužah in pri Žagi nad Kobaridom. Pod Kobaridom se dolina razširi (sl. 1.2,1.3). Med Kobaridom in Mostom na Soči jo je izoblikoval ledenik beds into the predominantly carbonate base. Dachstein limestone - which most of the Julian Alps consist of -prevails. In between we can find strips of dolomite, red marl and flysch, while the bottom of the valley is covered mostly by alluvial deposits. As far as the settlement of Žaga the Soča river runs parallel to the transverse Di-naric fractures, then it turns southwest and runs in the 'Dinaric' direction along the edge of the mountains of Krn. The valley narrows down whenever the river crosses a Dinaric ridge, e.g. at Kluže, or at Žaga above Kobarid. Sl. 1.2: Kobarid s Polovnikom in Kaninom v ozadju. Nad kostnico na Gradiču je Tonovcov grad. Pogled z juga. Fig. 1.2: Kobarid with Mt. Polovnik and Mt. Kanin in the background. Tonovcov grad stands above the ossuary on Gradič. na manj odpornj geološki podlagi. Sledi ledeniškega delovanja so vidne čelne morene pri Mostu na Soči. Dno doline predstavljajo ledeniško-rečne prodne terase, ki jih na zahodu omejujejo predalpski vrhovi Stol, Mija, Matajur in Kolovrat (Kunaver 1998, 54-70; Buser et al. 1989, 195-203). V Zgornjem Posočju se meša več fitogeografskih območij: alpsko, predalpsko in submediteransko-pred-alpsko. V dolinah prevladujejo bukovi gozdovi, pogosto pomešani s toploljubnimi listavci (črni gaber, mali jesen). Smreka raste predvsem na mestih kotanjastih površinskih oblik z zastajajočim hladnim zrakom. Višje na gorskih pobočjih se bukvi pridruži macesen, še višje pa ruševje (Dakskobler 1995, 167; Zupančič, Wraber 1989, 118-120). Delež gozdov v Zgornjem Posočju se v zadnjem stoletju povečuje predvsem zaradi zaraščanja pašniških površin. Arheološko najdišče Tonovcov grad je nizka (416 m) skalna gmota, ki je zadnji odrastek masiva 772 m visoke Babe v pogorju Kobariškega Stola (1300 m). Hrib leži na desnem bregu Soče tik nad reko, tako da popolnoma zapira prehod skozi sotesko. Vrh Tonovcovega gradu sega približno 200 m nad Sočo. Soteska je težko prehodna tudi na levem bregu, saj se tu nad Sočo dvigujejo pobočja Drežniške planote in Krna (sl. 1.2). Once it passes Kobarid the valley stretches out (Figs. 1.2, 1.3). Between Kobarid and Most na Soči the valley was formed by a glacier that slid across a less resistant geological base. The traces of the glacier can be seen in the lateral moraines at Most na Soči. The bottom of the valley is represented by terraces formed by glacio-fluvial gravel. In the west these terraces are bordered by the pre-alpine peaks of Stol, Mija, Matajur and Kolovrat (Kunaver 1998, 54-70; Buser et al. 1989, 195-203). A number of different phytogeographic landscapes meet in the Upper Soča Valley: Alpine, Pre-Alpine and Sub-Mediterranean-Pre-Alpine. Beech forests, often mixed with thermophilous deciduous trees (Hop Hornbeam, Manna Ash) predominate in the valleys. Spruce trees grow mainly in surface depressions where the air remains cold. Higher on the mountain slopes larch trees start appearing amongst the beech trees, and even higher up the dwarf pine (Dakskobler 1995, 167; Zupančič, Wraber 1989, 118-120). As a consequence of the grazing lands being abandoned and overgrown over the last century, the share of forests in the Upper Soča Valley has been on the increase. The archaeological site of Tonovcov grad is a low (416 m) rock formation, the last part of the 772 m high massif of Baba located in the mountain range of the Sl. 1.3: Položaj Tonovcovega gradu v soteski za Kobaridom tik nad Sočo. Pogled s severovzhoda. Fig. 1.3: The position of Tonovcov grad in the gorge behind Kobarid, above the Soča river. A view from the northeast. 1.1 GEOGRAFSKI ORIS 1.1 geographic description Sl. 1.4: Tonovcov grad. Pogled na najdišče z južne strani. Fig. 1.4: Tonovcov grad. A view from the south. Pobočja hriba, ki se z vseh strani strmo dvigujejo proti vrhu, so danes večinoma prerasla z gozdom, v katerem prevladujeta bukev in gaber (sl. 1.3, 1.4). Kobariški Stol (1300 m). The hill lies on the right bank of the river Soča, right above the river, so that it completely blocks the way through the gorge. The top of Tonovcov grad stands approximately 200 m above the river. On the left bank where the slopes of the Drežnica plateau and Krn rise above the Soča river the gorge is also difficult to pass (Fig. 1.2). The hill slopes that rise steeply towards the peak are nowadays mainly overgrown by forests, predominantly beech and hornbeam (Figs. 1.3, 1.4). 1.2 ime in izročilo 1.2 name and tradition Poleg v literaturi uveljavljenega in med ljudmi najbolje poznanega imena Tonovcov grad (domačini izgovarjajo Tonecov oziroma Tu n cov) je poznan tudi toponim Cekinov grad, ki je - kot zgovorno kaže navedba v zapuščini notarja Miroslava Premrova - odraz najdbe (najdb?) zlatnikov v preteklosti.1 Na nekaterih specialnih kartah je označen tudi kot Stari grad (npr. Atlas Slovenije 1986, 77). Prevladuje prvo ime, ki smo ga tako, kot ga je v prvi tiskani omembi zapisal zgodovinar Simon Rutar, sprejeli tudi arheologi, saj povsem jasno razlikuje to arheološko najdišče od številnih "Cekinovih" in "Starih" gradov po Sloveniji. Tudi domačini danes obeh zadnjih imen ne uporabljajo, oziroma zanje v večini ne vedo več.2 Da je imel Tonovcov grad v zavesti okoličanov pomembno mesto, opozarja bogato izročilo: razširjena je pripoved o gradu na njem (glej označbo Stari grad na specialnih zemljevidih), o njegovih zakladih in roparskih vitezih, ki so ropali potnike na cesti pod gradom. V pesniško oblikovani pripovedi o Livškem jezeru je Anton Klodič - Sabladoski zapustil celo dokaj podroben opis roparskega gradu nad Sočo, kjer omenja močno obzidje in štiri stolpe (Klodič - Sabladoski 1912). Pripoved o treh bratih, roparskih vitezih, ki so prežali na popotnike in tovornike, pa je zapisana v zbirki pripovedk (Dolenc 1992, 159). V ustnem izročilu so znane tudi zgodbe o prostorni jami pod najdiščem, na kar se morda nanaša izročilo o "treh tinah zlata v tretji kleti" pod Tonovcovim gradom.3 Rimsko cesto pod Tonovcovim gradom omenja že Simon Rutar (Rutar 1890, 129). Cesta naj bi bila speljana čez hribček Sv. Antona in od tam vodila polagoma navzdol v globel za Tonovcovim gradom, kjer je domneval obstoj srednjeveškega gradiča Pottenstein. Tu naj bi se še dobro videl cestni tlak in kolesnice. Njen potek naj bi se kmalu za Tonovcovim gradom združil z današnjo Apart from the name Tonovcov grad (pronounced by the locals Tonecov) which is used in literature and best known amongst the locals, the settlement is also known as Cekinov grad (Sequin castle), which is - as indicated by a reference in the legacy of the notary Miroslav Premrov - a result of the find (finds?) of gold coins in the past.1 On some survey maps the site is also marked under the name of Stari grad (Old Castle, e.g. Atlas Slovenije 1986, 77). The first name is the most common, and archaeologists have adopted it in the way it was used by historian Simon Rutar when it first appeared in print, for it clearly separates this archaeological site from the numerous 'Gold' and 'Old' castles around Slovenia. The locals have stopped using the latter two names, and in most cases they do not even know about their existence.2 The rich oral tradition indicates that Tonovcov grad always held an important position in the consciousness of the locals: there is the story of a castle on the hill (see the name Stari grad on survey maps), and the stories about treasures and robber knights who robbed the travellers on the road beneath the castle. In the poem by Anton Klodič - Sabladoski about the Livek lake the author even left behind quite a detailed description of the robber castle above the Soča river, in which he mentions strong fortification walls and four towers (Klodič - Sabladoski 1912). A collection of stories (Dolenc 1992, 159) includes A tale of three brothers - robber knights - who wait for travellers below the castle. The oral tradition also includes stories describing a large cave under the castle. Perhaps they are the source of the 'three units of gold in the third basement' under Tonovcov grad.3 The Roman road below Tonovcov grad was first mentioned by Simon Rutar (Rutar 1890, 129). Supposedly it led over the small hill of Sv. Anton from where it descended slowly into the hollow behind Tonovcov grad, 1 Zapuščina notarja Premrova. Hrani arhiv Narodnega muzeja Slovenije v Ljubljani. 2 Pismo Meri Koren z dne 5. 6. 1993. Hrani Inštitut za arheologijo ZRC SAZU. 3 Podatek v pismu Meri Koren z dne 27. 6. 1994. Hrani Inštitut za arheologijo ZRC SAZU. 1 The legacy of notary Premrov. Kept in the archive of the Narodni muzej Slovenije in Ljubljana. 2 Letter from Meri Koren, dated 5th June 1993. Kept at the Institute of Archaeology at ZRC SAZU. 3 Information from the letter by Meri Koren, dated 27th June 1994. Kept at the Institute of Archaeology at ZRC SAZU. cesto proti Trnovem. Arheološke raziskave ceste še niso bile opravljene, je pa glede na konfiguracijo terena zelo verjetno tekla po trasi, ki jo je opisal že Rutar. Nedvomno cesta zaradi ugodne lege na sicer zahtevnem terenu utegne biti naslednica v večjem delu starih poti, če že ne "prave" rimske ceste (glej tudi pogl. 1.6). Izročilo o "rimski cesti", ki je vodila od Gradiča mimo Tonovcovega gradu naprej proti Trnovem je še vedno živo. Domačin Zdravko Likar se spominja tudi večjega kamnitega valja, ki ga je imenoval "miljnik" in naj bi še pred desetletji ležal ob cesti v predelu takoj za Gradičem. Kamna danes na terenu ni več mogoče izslediti. where he assumed the medieval castle of Pottenstein existed. The road pavement and the wheel ruts were supposed to be still visible. After it passed Tonovcov grad it supposedly joined the present-day road towards Trnovo. So far archaeological research of the road has not been carried out, but according to the configuration of the terrain it is highly likely that it did run along the route described by Rutar. Regarding its favourable position on an otherwise difficult terrain it is probable that the major part of the current road runs along the old routes if not the 'real' Roman road (see also chapter 1.6). The oral tradition about the 'Roman road' that lead from Gradič past Tonovcov grad and towards Trnovo is still alive. The local Zdravko Likar remembers that a large cylindrical stone that he took for a milestone lay at the roadside just behind Gradič even a few decades ago. Nowadays the stone could not be located anymore. 1.3 opis najdišča 1.3 site description Hrib Tonovcov grad z večplastnim arheološkim najdiščem na vrhu meri v dolžino približno tristo metrov. Za naselbino je bil uporabljen le najvišji, naravno najugodneje zavarovan zahodni del hriba, medtem ko je vzhodna polovica v zgornjem delu nekoliko zložnejša in se šele v spodnjem delu prevesi v strmo skalnato steno. Najlažji je dostop s severne strani. Sedanja pot je lepo speljana in se enakomerno vzpenja po severni strani po- The hill of Tonovcov grad and the multilayer archaeological site on its top measure approximately 300 meters in length. Only the highest, naturally best protected west side of the hill was used for the settlement. The east side is less sloping at the top and it only turns into a steep rock face in its lower part. The easiest access is from the north. The current path runs smoothly and climbs at an equal pace along the north slope. On a smal saddle it turns Sl. 1.5: Tonovcov grad z okolico. Lidarski posnetek (© ZRC SAZU). Označena današnja (domnevno poznoantična) pot, današnja cesta in reka Soča (skrajno desno). Fig. 1.5: Tonovcov grad and its surroundings on a lidar data visualization (© ZRC SAZU). The modern (supposed Late Antique) path and the modern road are marked. The Soča river is to the far right. Sl. 1.6: Tonovcov grad. Na površini strmega pobočja so vidni ostanki ruševine prečnega zidu. Fig. 1.6: Tonovcov grad. The remains of the transverse wall as seen on the surface of the steep slope. bočja. Na manjšem sedlu zavije proti vzhodu ter pripelje do manjšega, deloma umetno zravnanega prostora (sl. 1.5,1.7). Tega je v poznoantičnem času branil prečni zid na zahodni strani severnega pobočja, katerega ruševino je mogoče slediti po strmini vse do vrha hriba, kjer se je priključil k obrambnemu zidu zgornjega dela naselbine (sl. 1.6, 1.7). Prečni zid je predstavljal prvo obrambno linijo na najlažje dostopni strani. tako omejen prostor ima na severni strani dokaj ostro zaključen rob, ki utegne towards the east and leads to a smaller, partly artificially levelled area (Figs. 1.5, 1.7). In Late Antiquity this area was defended by a transverse wall on the west side of the northern slope, the ruins of which can be traced all the way to the top of the hill, where it joined the defensive wall of the upper part of the settlement (Figs. 1.6, 1.7). This transverse wall represented the first line of defence on the most accessible side. This limited area ends on the north side with a relatively sharp edge, which could be Sl. 1.7: Tonovcov grad. Načrt najdišča z raziskanimi (št. 1-5) in v reliefu površine nakazanimi stavbami (št. 6-31). Fig. 1.7: Tonovcov grad. Site plan of the settlement with the researched buildings (Nos. 1-5) and the buildings that can be seen in the surface (Nos. 6-31). skrivati obrambni zid: na njegovem robu je videti sledove stavbe, ki je brez raziskav ni mogoče zanesljivo datirati (morda tudi ostanki iz časa prve svetovne vojne?). Pot se od sedla čez pobočje zlagoma dviga proti naselbini. Speljana je bila tako, da so morebitni napadalci prihajali po njej obrnjeni proti obrambnemu zidu z desno stranjo, ki ni bila zavarovana. Prav zaradi tako speljane poti in jasno razpoznavne vrzeli v sicer strnjeni vrsti stavb ob obzidju sklepamo, da današnja pot v celoti vodi po trasi prvotnega pristopa. Mesto naselbine je bilo izbrano izredno premišljeno: večji del stavb leži v nekoliko uleknjenem delu temena hriba, ki se na vzhodnem in zahodnem robu postopno dvigne za nekaj metrov. Teren se postopoma dviga tudi na južni strani, tako da je bilo jedro naselbine pred vetrovi odlično zaščiteno kar s treh strani (sl. 1.8). utrjena naselbina na vrhu hriba obsega predvsem zahodni del skalne kope. S strminami in obzidjem obdani prostor meri 150 x 90 m. Obrambni zid debeline 80 cm je viden na vzhodni strani, kjer ločuje naselbino od hiding a defensive wall. On the edge traces of a building are visible, but without further research they cannot be reliably dated (possibly World War I remains?). The path rises gradually from the saddle towards the settlement. It was made in such a way that the eventual attackers would have to approach the defensive wall with their undefended right side exposed. This and the clearly recognisable gap in the otherwise solid line of buildings along the wall led us to assume that today's path exactly follows the original line of approach. The settlement location was carefully selected: the majority of the buildings lie in the slightly lower part of the ridge, which rises gradually (by a few meters in the east and in the west). The terrain also rises gradually to the south, thus the core of the settlement was well protected from the winds on three sides (Fig. 1.8). The fortified settlement on the top of the hill covers above all the western part of the rock. The area is surrounded by steep slopes and the defensive wall and measures 150 x 90 m. The 80 cm thick defensive wall Sl. 1.8: Tonovcov grad. Pogled s severozahodne strani na osrednji del naselbine z izkopano stavbo št. 1 v ospredju. Fig. 1.8: Tonovcov grad. A view of the central part of the settlement from the northwest, with the excavated building No. 1 in the foreground. nekoliko nižjega vzhodnega dela hriba. Brez izkopavanj ni mogoče zanesljivo reči, ali je bil zid postavljen na kakšnem zemljenem - morda prazgodovinskem - nasipu ali pa je na terenu vidni 1-3 m visok nasip le ruševina nekdaj veliko višjega obzidja (sl. 1.9). Nadaljevanje zidu proti jugu ni ohranjeno ali pa obzidje ni bilo zgrajeno povsem do roba južne skalne stene. Morda bi lahko tu domnevali drugi vhod v naselbino, saj je bila potrebna can be seen on the east side, where it separates the settlement from the slightly lower eastern part of the hill. Without excavations it is impossible to say whether the wall stood on a - maybe prehistoric - rampart or whether the 1-3 m high dike is merely the ruin of the once much higher wall (Fig. 1.9). Its continuation towards the south is not preserved or the wall was not built all the way to the edge of the south rock face. Perhaps this was the second Sl. 1.9: Tonovcov grad. Ruševinski nasip z delno vidnim obzidjem na vrhu in ruševino stolpa (št. 31) na vzhodni strani naselbine. Fig. 1.9: Tonovcov grad. A dike with partially visible wall at the top and the tower ruins (No. 31) on the east side of the settlement. Sl. 1.10: Tonovcov grad. Pogled z vzhodne strani na ruševino stavbe št. 28 po viharju pozimi leta 2008. Fig. 1.10: Tonovcov grad. A view of the ruins of building No. 28 soon after the storm in winter 2008, from the east. komunikacija s preostalim, manj zavarovanim delom hriba. Obzidje je slabše vidno na severni strani, saj so bile tu v večjem delu zgrajene manjše stavbe, tesno razvrščene druga poleg druge in prislonjene na obzidje. Ni mogoče zanesljivo razločiti, ali gre za sklenjeno vrsto stavb ali pa so bile te na posameznih mestih ločene z vmesnimi prostori. v notranjosti je bilo mogoče po temeljitem čiščenju podrasti opaziti ruševine stavb, katerih tlorisi so v površini tako dobro vidni, da je mogoče dokaj natančno rekonstruirati nekdanjo naselbino, pri večini stavb pa celo ločiti število prostorov in razbrati njihov razpored (sl. 1.7). Kot smo že omenili, so najpreprostejše stavbe, ki ležijo ob severnem obrambnem zidu. Izrazitejše so tiste znotraj zavarovanega območja. Izstopa zgradba, ki leži povsem blizu vhoda in nadzira vstop v naselbino: zavzemala je osrednji prostor v naselbini in hkrati ležala na dokaj ravnem terenu - pri izkopavanju smo jo označili kot stavbo 1 (sl. 1.7; glej pogl. 2.3). Dve triprostorni stavbi (objekta št. 22 in 26, sl. 1.7) ležita v neposredni bližini: pri eni od njiju (objekt št. 22) se zdi, da je bil osrednji prostor močneje grajen ali vsaj višji, saj kot ruševina izraziteje izstopa iz talnega reliefa. Še ena triprostorna stavba (v kolikor ne gre za tri samostojne objekte) je stala tik pod vrhom zahodnega zaključka naselbine, na manjši, lepo položeni terasi, vendar je entrance to the settlement as communication with the other, less protected part of the hill, was also necessary. The defensive wall is less clearly visible on the north side where small buildings were built close together and leaning upon the wall. It is difficult to distinguish whether this was a line of attached buildings or were they separated by empty spaces. After the undergrowth in the interior was cleared we discovered the ruins of buildings, the ground plans of which were so clearly visible on the surface that we could relatively precisely reconstruct the former settlement, and in most buildings we could even distinguish the number of rooms and their layout (Fig. 1.7). As already mentioned, the simplest buildings are positioned alongside the north defensive wall. The buildings within the protected area are more complex. Very conspicuous is a a building that lies very close to the entrance to the settlement and must have controlled it. This building occupied the central area within the settlement and stood on relatively level terrain - during the excavations it was named building 1 (Fig. 1.7; see chapter 2.3). Two three-room buildings (buildings No. 22 and 26, Fig. 1.7) can be found in the immediate vicinity. One of them (building No. 22) seems to have had a better built or at least a taller central space for its ruin is visibly higher than others. Another three-room building (or perhaps three individual buildings) Sl. 1.11: Tonovcov grad. Pogled z južne strani na v celoti izkopan sklop zgodnjekrščanskih cerkva. Fig. 1.11: Tonovcov grad. A view of the three excavated Early Christian churches, from the south. bila prilagojena terenu, saj si prostori v njej ne sledijo v povsem ravni črti (objekt št. 14, sl. 1.7). V severozahodnem vogalu naselbine je moral stati neki pomembnejši objekt, saj je bilo prav to mesto eno izmed ključnih delov obrambe celotne naselbine. Vogal obzidja je še prav dobro viden, nista pa jasni izoblikovanost in velikost stavbe oziroma stolpa, ki je stal na tem mestu (objekt št. 18, sl. 1.7). Zaradi nekoliko nagnjene lege se je ruševina nedvomno večinoma posula po pobočju. Kot posebnost izpostavljamo ruševino stavbe s petimi prostori (objekt št. 28, sl. 1.7), ki so razporejeni po vzhodnem pobočju na dveh višinskih nivojih (sl. 1.10). Na nekaj metrov visokem skalnem platoju, ki se na južni strani navezuje na južno skalno steno naselbine, je bil že pred izkopavanjem dobro viden največji ruševinski sklop (objekt št. 4, sl. 1.7). To so sledovi zgodnjekr-ščanskih cerkva, ki smo jih kot take prepoznali že pred sistematičnim izkopom (glej pogl. 2.5). Izpostavljeno lego na robu naselbine ima tudi stavba (objekt št. 8, sl. 1.7), ki leži nekako v sredini južnega roba, tik nad prepadno steno. Glede na dolžino in orientacijo se sklada s cerkvenimi zgradbami na nekoliko višje ležečem platoju (sl. 1.12). Tik pod njenim vzhodnim zaključkom je v manjšem sedlu vidna enopro-storna stavba, ki je veliko nižja in nekoliko poglobljena (objekt št. 7, sl. 1.7). V njej domnevamo prej omenjeni večji stavbi pripadajoči vodni zbiralnik. stood right under the top of the west end of the settlement on a small, well positioned terrace. However it was still adjusted to the terrain, for the rooms were not built in a straight line (building No. 14, Fig. 1.7). An important building had to stand in the northwest corner of the settlement, for this spot was one of the key points in the defence of the entire settlement. The corner of the fortification wall is still clearly visible, but the size and shape of the building or tower that stood there is not entirely clear (building No. 18, Fig. 1.7). Due to the slightly sloping terrain it can be assumed that most of the ruin slid down the slope. An interesting feature among the house layouts is the ruin of a building with five rooms (building No. 28, Fig. 1.7) which are positioned on two different height levels along the east slope (Fig. 1.10). A slightly elevated rock plateau joins the rock face in the southern part of the settlement. There the largest complex of ruins was visible (building No. 4, Fig. 1.7) even before the excavations started. These were remains of Early Christian churches that were recognised as such even before systematic research (see chapter 2.5). Another building (building No. 8, Fig. 1.7) situated roughly in the middle of the south edge of the settlement, just above the precipice, was also very exposed. In length and orientation it resembles the church buildings on the plateau (Fig. 1.12). In a small saddle just under its east Sl. 1.12: Tonovcov grad. Visoko ohranjena ruševina stavbe št. 8 na robu naselbine (domnevno cerkev) in pred njo globlji prostor stavbe št. 7 (domnevni vodni zbiralnik). Fig. 1.12: Tonovcov grad. The high preserved ruins of building No. 8 at the edge of the settlement (assumed to be a church) and the depressed area of building No. 7 (presumably a water reservoir) in front of it. Še en vodni zbiralnik (objekt št. 5, sl. 1.7) je bil raziskan na najvišjem, vzhodnem platoju naselbine. Vmes med omenjenimi večjimi zgradbami je še nekaj manjših, najverjetneje enoprostornih, ki dopolnjujejo podobo naselbine. Po neurju, ki je pozimi 2008 izruvalo in polomilo nekaj dreves, se je ob čiščenju terena pokazala še ena tlorisna zasnova z več prostori: leži povsem na robu platoja z vodnim zbiralnikom in je bila na prvotnem načrtu nakazana le kot manjša dvoprostorna zgradba zahodno od zbiralnika (objekt št. 30). Njen tloris pa bo v celoti mogoče vrisati šele z nadaljnjimi raziskovanji. Nedvomno je bil njen pomen precejšen, saj gre za najvišje ležečo veliko stavbo z dokaj razčlenjeno tlorisno zasnovo. Na hribu so tudi sledovi iz prve in druge svetovne vojne. Poleg stopnic in bunkerja na vzhodnem pobočju hriba so na vrhu še jarki, kaverna in večji vkop iz druge svetovne vojne, ki so deloma poškodovali poznoantično plast. Nemški bunker so v celoti vgradili v poznoantični vodni zbiralnik (objekt št. 5) in tako izkoristili dokaj dobro ohranjeno antično strukturo (glej pogl. 2.6). end a single room building, much lower and slightly deepened, is visible (building No. 7, Fig. 1.7). Presumably this is a water reservoir that supplied the larger building. Another water reservoir (building No. 5, Fig. 1.7) was researched on the highest, eastern plateau. In between the mentioned larger buildings there are a few smaller, most likely single-room buildings that complete the layout of the settlement. After the terrain was cleared from the storm that toppled and snapped a number of trees in the winter of 2008 another building with several rooms was revealed. It is situated on the very edge of the plateau with the reservoir and was previously thought to be merely a smaller two-room building to the west of the water reservoir (building No. 30). Further research will be needed to reconstruct its ground plan. It was undoubtedly important, for is situated on the highest part of the settlement and has a relatively complex ground plan. The hill also shows traces from the two world wars. Apart from the stairs and bunker on the east slope of the hill, some ditches, a cavern and a large trench from World War II on the top of it partially damaged the late Antique layer. Making use of the relatively well preserved Antique structures the German bunker was built within the late Antique water reservoir (building No. 5; see chapter 2.6). 1.4 zgodovina raziskav 1.4 research history Posočje v celoti, posebej pa še njegov severni del, je bilo arheološko dolgo časa slabo znano. Pozornost raziskovalcev je bila usmerjena predvsem v veliko in izjemno pomembno prazgodovinsko postojanko v Mostu na Soči (Sv. Lucija), kjer so že v 19. st. izkopali grobišče, po drugi svetovni vojni pa so pri zaščitnih raziskavah v samem jedru kraja odkrili odlično ohranjene sledove prazgodovinskih stavb (pregled raziskav pri Mlinar 2002). Vzporedno s temi deli so potekala zgolj manjša zaščitna izkopavanja, opravljeni so bili redki topografski pregledi ter odkrite posamezne naključne najdbe arheoloških predmetov (ANSl 1975, 115-119; 124-127). Tudi ožje kobariško območje je bilo dolgo deležno enake usode raziskanosti kot celotno Posočje. Kljub izkopu druge največje halštatske nekropole na pobočjih in ravnici med Sočo in Gradičem že ob koncu 19. st. je velika večina tukajšnjega gradiva obležala neobjavljena v depojih, nekropoli pripadajoče gradišče pa je bilo povsem pozabljeno (ANSl 1975, 116). Večje zaščitne raziskave grobišča je tu leta 1979 opravil Drago Svoljšak (Svoljšak 1981). Šele z odkritjem antičnih bronastih plastik spodbujena zaščitna izkopavanja Zavoda za varstvo naravne in kulturne dediščine pod vodstvom Nade Osmuk na pobočju Gradiča so znova opozorila na velik arheološki pomen ravnice med Sočo in Nadižo, na izjemen prometni in strateški pomen tega prostora v prazgodovinskem in rimskem obdobju kot tudi na staro kultno središče na tem območju (Osmuk 1987a; 1997a). Poznoantične in zgodnjesrednjeveške najdbe so bile neznane, še tistih nekaj elementov, ki so kazali na obstoj najdišč, je bilo spregledanih ali pozabljenih. Delno je to vrzel leta 1991 zapolnil Matej Župančič, ki je opozoril na starejšo najdbo zlatega solida iz sredine 5. st. z Molide, in nakazal možnost obstoja poznoantične postojanke (Župančič 1991). Arheološko najdišče Tonovcov grad pa je imelo povsem samosvojo, nenavadno usodo in bilo kljub nekaterim zgodnejšim omembam in starejšim naključnim najdbam zelo pozno prepoznano kot eno izmed najcelovitejših večplastnih arheoloških najdišč z izjemno ohranjeno poznoantično naselbino. The archaeology of the Soča Valley, especially its north part, was for a long time very poorly known. The attention of the researchers was predominantly directed towards the large and extremely important prehistoric settlement at Most na Soči (Santa Lucia), where a cemetery was excavated as early as the 19th century. Rescue excavations after World War II revealed excellently preserved traces of prehistoric buildings in the very centre of the present-day settlement (for an overview of the various researches see Mlinar 2002). Apart from that there were only some minor protective excavations, rare surveys and individual archaeological finds (ANSI 1975, 115-119; 124-127). The Kobarid area suffered similar fate. Even though the second largest Hallstatt necropolis in Slovenia was discovered on the slopes between the Soča river and Gradič as early as at the end of the 19th century a great portion of the excavated material remained unpublished, and the hillfort that belonged to the necropolis was completely forgotten (ANSI 1975, 116). A protective excavation of the necropolis took place in 1979 and was led by Drago Svoljšak (Svoljšak 1981). It was only with the discovery of the Antique bronze statues that rescue excavations on the slope of Gradič were started by Nada Osmuk from the Institute for the Protection of the Natural and Cultural Heritage. This discovery brought back the attention to the great archaeological importance of the straights between the Soča and Nadiža rivers, its exceptional strategic and transport importance during the prehistoric and Roman period as well as to the old cult centre in this area (Osmuk 1987a; 1997a). Late Antique and Early Medieval finds were unknown, even the few elements that indicated the existence of sites, were overlooked or forgotten. This gap was partially filled in 1991 by Matej Župančič, who drew attention to an old find of a mid 5th century gold solidus from Molida, that indicated a possibility of the existence of a Late Antique post (Župančič 1991). The archaeological site of Tonovcov grad had a unique and unusual fate. Despite some early references and accidental finds it was very late to be recognised as one of the most complex multi-layered archaeologi- Najdišče je bilo v arheološki literaturi prvič omenjeno leta 1882, ko ga je zgodovinar Simon Rutar, tolminski rojak, v svoji Zgodovini Tolminskega navedel kot možno arheološko lokacijo (Rutar 1882, 12; glej pogl. 1.2). Sam si ga sicer ni ogledal, navaja pa pripovedi, ki omenjajo razvaline gradu. Očitno tem pripovedim ni zaupal, saj nakazuje tudi možnost, da je "stari grad" dobil ime po obliki hriba, kar naj bi se pogosto dogajalo. V poročilu o prazgodovinskih in rimskih izkopavanjih na Slovenskem se je leta 1890 ponovno dotaknil tega najdišča (Rutar 1890, 129). Omenja ga v zvezi z rimsko cesto, ki naj bi tekla od hribčka Sv. Antona polagoma navzdol v jarek za "Tonovčevim gradom". Pri tem domneva, da naj bi tu v srednjem veku stal gradič "Pottenstein". Nekaj podatkov o Tonovcovem gradu se je ohranilo tudi v rokopisni zapuščini M. Premrova, notarja v Kobaridu, ki v svoji "Gedenkbuch der Pfarr Karfreit" z dne 24. 2. 1915 omenja kopanje na hribu "Tonovcev gradc", kjer na bi pred pribl. 50-60 leti (torej okoli 1860) šest domačinov, z očetom Tonovcom vred, kopalo jamo, v kateri so zadeli na zidove.4 Omenja tudi govorico, po kateri naj bi fant iz Ladrij našel na krtini cekin. Kljub tem navedbam najdišče med strokovnjaki ni izzvalo večjega zanimanja, zato ga pri pregledu Kobarida in okolice ne upošteva niti temeljni pregled slovenskih arheoloških najdišč Arheološka najdišča Slovenije iz leta 1975 (ANSl 1975, 116). Leta 1985 je hrib Tonovcov grad obiskala arheologinja goriškega Zavoda za varstvo naravne in kulturne dediščine Nada Osmuk, ki je opazila pomembno strateško lego in odlično obrambno naravo najdišča (Osmuk 1985c). Opozorila je na sledove obrambnega okopa na jugovzhodni in južni strani, sledov kulturne plasti pa na površini ni ugotovila. Časovno opredelitev je zaradi pomanjkanja izpovednejših najdb (omenja le opeke) pustila odprto. Izrazila je mnenje, da je domnevna utrjena postojanka zapirala dostop v kobariško kotlino s severa ob cesti na desnem bregu Soče, in ugotovila, da ima postojanka ugodnejši strateški položaj od Gradiča. Na časovno uvrstitev utrjene naselbine pa so opozorile najdbe nekaterih izpovednih arheoloških predmetov iz vrhnje plasti Tonovcovega gradu, ki sta jih našla Žiga Šmit in Tomislav Drčar.5 Prve najdbe je leta 4 Hrani arhiv Narodnega muzeja Slovenije v Ljubljani. Podatek mi je prijazno posredoval Silvo Torkar z Inštituta za slovenski jezik Frana Ramovša ZRC SAZU. 5 Takratne najdbe sestavljajo skupino predmetov, odkritih po celotni površini najdišča in po pobočjih zunaj njega, ki so bili izkopani v zgornji plasti, v ruši, in so, kot smo ugotovili pozneje pri sistematičnih izkopavanjih, ležali v plasti, ki je bila poškodovana z gostim koreninskim prepletom vegetacije, ki je v zadnjih sto letih zarasla prej neporaščeni To-novcov grad. Ti predmeti so brez stratigrafskih kontekstov, ki jih v gozdni površinski plasti - pretežno v območjih med stavbami - tudi ni mogoče pričakovati. Izjema so le ostanki velike bronaste posode (Tonovcov grad. Najdbe, t. 50: 1), ki je bila najdena v ruševinski plasti narteksa južne cerkve, kot cal sites with an exceptionally preserved Late Antique settlement. The settlement was first mentioned in archaeological literature in 1882, when the historian Simon Rutar from Tolmin, mentioned it as a possible archaeological location in his book Zgodovina Tolminskega [History of the Tolmin area] (Rutar 1882, 12; see chapter 1.2). He did not visit it himself, but writes about stories that mention castle ruins. He obviously did not find these stories reliable, for he also indicated the possibility that the 'old castle' obtained its name from the shape of the hill (which was apparently a rather common occurrence). He mentioned the site again in 1890 in his report on the Prehistoric and Roman excavations in Slovenia (Rutar 1890, 129): a Roman road was supposed to have ran from the hill of Sv. Anton down to the hollow behind 'Tonovčev grad' and the small castle of 'Pottenstein' was supposed to have stood on the hill in the Middle Ages. Some data on Tonovcov grad was preserved in the manuscripts of M. Premrov, a notary in Kobarid, who in his 'Gedenkbuch der Pfarr Karfreit dated on 24th February 1915 mentions digging that took place on the hill 'Tonovcev gradc, when approximately 50-60 years ago (i.e. around 1860) six locals, including the owner ('father Tonovc'), discovered walls while digging a pit.4 He also mentions a rumour according to which a boy from the village Ladra found a gold coin in a molehill. Regardless of this the site did not raise interest amongst experts, even the major overview of Slovenian archaeological sites Arheološka najdišča Slovenije from 1975 (ANSl 1975, 116) fails to mention it in the overview of Kobarid and its surroundings. In 1985 the hill Tonovcov grad was visited by Nada Osmuk, an archaeologist from the Gorica Institute for the Protection of National and Cultural Heritage. She noticed the important strategic position and the excellent defensive nature of the site (Osmuk 1985c). She drew attention to the traces of the rampart on the southeast and south side, however she failed to discover any traces of the cultural layer on the surface. Due to the lack of any revealing finds (she only mentions tegulae) she could not date the site. She noted that the assumed fortified post closed off the north entrance into the Kobarid basin on the right bank of the river Soča, and that it had a better strategic position than Gradič. The chronology of the fortified settlement was implied by some revealing artefacts discovered by Žiga Šmit and Tomislav Drčar in the topmost layer of Tonovcov grad.5 The first finds were gathered in 1991 by Ž. Šmit 4 Kept in the archive of the Narodni muzej Slovenije in Ljubljana. This information was kindly passed onto me by Silvo Torkar from the Fran Ramovš Institute of the Slovenian Language at ZRC SAZu. 5 The finds include a group of objects that were discovered across the entire surface of the site and along the slopes outside of it, all of which were discovered in the top turf layer, 1991 pridobil Ž. Šmit z iskalcem kovin in jih 20. 9. 1991 prinesel na Inštitut za arheologijo ZRC SAZU. Sodelavci Inštituta smo se zato takoj odpravili na najdišče, da bi preverili mesto najdb. Po analogiji z že prej odkritim Gradcem pri Prapretnem (Ciglenečki 1975, 259-267) smo odkrili številne na površini v reliefu dobro vidne stavbne ostaline iz poznoantičnega obdobja. Datacijo stavb sta potrjevali predvsem gradnja zidov in sestava malte v zidu velike stavbe v bližini vhoda, ki je bil prekopan že pred davnimi leti - najverjetneje, kot omenja Premrov, so tu kopali lastnik Tonovc in njegovi tovariši. Po številnih terenskih obhodih najdišča in okolice smo spomladi 1992 pripravili tudi natančen geodetski posnetek, v katerega smo vrisali vse na površini vidne strukture in vnesli vanj mesta izpovednejših najdb, ki sta jih našla Ž. Šmit in pozneje T. Drčar.6 Najdišče je bilo v smeri S-J razdeljeno na kvadratno mrežo z velikostjo kvadrantov 4 x 4 m, obenem pa so Ž. Šmit, T. Drčar in Zoran Božič natančno preiskali površino. Pri tem smo ravnali skladno z navodili, ki jih je objavila spomeniška služba (Curk 1994/1995). Tako nam je uspelo vse do takrat odkrite najdbe vnesti v geodetski posnetek najdišča in okolice, s čimer smo omogočili poznejšo rekonstrukcijo eventualnih zaključenih sklopov. Ob tem je treba poudariti, da je bila velika večina tako pridobljenih arheoloških predmetov odkrita do globine 15 cm in v arheološko ne več intaktnih plasteh. Zaradi pomembnosti izjemno ohranjenih struktur poznoantične naselbine smo pripravili načrt raziskovanj, da bi odlično ohranjeno, a slabo znano naselbino čim prej vključili v sistematična raziskovanja poznoantičnega obdobja Slovenije. Glede na do tedaj znano tipološko klasifikacijo poznoantičnih višinskih postojank je prav ta naselbina kazala najkompleksnejši značaj in posrečeno združevala kar nekaj najbolj tipičnih elementov poznoantičnih utrdb: veliko in dobro ohranjeno naselbino, katere obrisi so bili dobro vidni na površini, pomembno strateško lego, na zgornjem platoju naselbine pa se je v reliefu nakazovalo veliko zgodnjekrščansko središče z več cerkvami. Nekoliko senzacionalističen opis ("slovenski Pompeji") navdušenega obiskovalca raziskav na Tonovcovem gradu in ljubitelja arheologije Zorana Božiča v Primorskih novicah je sprožil veliko zanimanje za najdišče, hkrati pa izzval odgovor tedanjega kobariškega dekana Franca Rupnika (Božič 1993; Rupnik 1993). Čeprav so bile raziskave šele v začetni fazi, je zato moral slediti še odgovor pristojnega arheologa, da se "novi kobariški čudež" in "slovenski Pompeji" ne bi preveč neverodo- with the use of a metal detector and brought to the Institute of Archaeology at ZRC SAZU on 20. 9. 1991. With the co-workers from the Institute we immediately set off to the site in order to check the location of the finds. Comparing the surface shape of the site with the recently discovered Gradec near Prapretno (Ciglenečki 1975, 259-267) numerous building remains from the Late Antique period were found out. The dating of the buildings was confirmed by the construction of the walls and mortar composition on the wall of the big building close to the entrance. This building has been previously damaged by a robber trench - probably this was the location where the owner Tonovc and his companions had dug as mentioned by Premrov. In the spring of 1992, after numerous field inspections of the site and its surroundings we prepared a precise land survey record, into which we entered all structures that could be seen on the surface and the locations of the more revealing finds discovered by Ž. Šmit and later T. Drčar.6 The site was divided into a square grid running in the north -south direction, with the quadrants measuring 4 x 4 m. Ž. Šmit, T. Drčar and Zoran Božič inspected the topsoil in great detail. They proceeded in accordance with the instructions published by the office for protection of cultural heritage (Curk 1994/1995). We managed to include all of the discovered finds in the survey map of the site and its surroundings, so as to provide a good basis for eventual later reconstructions. Here it should be emphasised that the majority of the artefacts obtained in this manner were discovered up to 15 cm deep in layers that were not archeologically intact. Because of the importance of the exceptionally well preserved structures of Late Antique settlement we prepared an excavation plan that would enable us to include the excellently preserved, but poorly known site into the systematic research of Late Antiquity in Slovenia as soon as possible. In the typology of the Late Antique hilltop posts this settlement showed an extremely complex character and combined some of the most typical elements of Late Antique forts: a large and well preserved settlement, clearly visible in the surface shape, a strategically se je izkazalo pri izkopavanju leta 1996. Lego predmetov smo skupaj z najditeljema natančno vnesli v geodetski posnetek naselbine, da bi ob poznejšem izkopavanju pomagali pri tolmačenju najdiščnih sklopov. 6 Geodetsko izmero je opravil geodet Bogo Žontar (t 2003), ki je z veliko prizadevnostjo in veseljem spremljal naša izkopavanja. that was, as we have ascertained during the systematic excavations, damaged by the roots of the dense vegetation which has overgrown the previously bare Tonovcov grad during the last one hundred years. These objects have no stratigraphical contexts, which is of course to be expected in a forest surface layer - mainly in the areas between the buildings. The exceptions are the remains of the large bronze vessel (Tonovcov grad. Finds, Pl. 50: 1) that was discovered in the destruction layer of the narthex of the south church as was established during the excavations in 1996. Together with the two finders we entered the precise locations of the objects into the geodetic map of the settlement, hoping that this would help explain the site during the later excavations. 6 The geodetic measurements were made by the land surveyor Bogo 2ontar (t 2003), who followed our excavations with great dilligence and enthusiasm. stojno zasidrali v zavesti slovenskih ljubiteljev antične zgodovine. Slavko Ciglenečki je na podlagi do tedaj znanih izsledkov pojasnil osnovne značilnosti novoodkrite poznoantične postojanke in tako odgovoril na dileme, ki so se o značaju najdišča porodile v časopisnih objavah (Ciglenečki 1993). Leta 1993 smo opravili prvi raziskovalni poseg v naselbini. Osredotočili smo se na saniranje že izkopane večje jame na zgornjem platoju naselbine (med osrednjo in južno cerkvijo), očistili že močno porušene robove izkopa in tako pridobili prvi uvid v kulturne plasti najdišča.7 Poleg ruševin gradbenih ostankov smo izkopali večje število odlomkov poznoantične keramike (grobe kuhinjske lončenine, amfor in afriške sigilate) in antične kritine (imbreksov in tegul). Prve podrobnejše strokovne rezultate preliminarnih raziskav smo objavili leta 1993 v reviji Kronika, kjer je bila tematska številka posvečena prav zgodovini Tolminskega (Ciglenečki 1994b). Nekoliko pozneje smo neznatno dopolnjeno verzijo prevedenega besedila objavili tudi za italijansko strokovno javnost (Ciglenečki 1994c). Začetne raziskave so tako okvirno določile razpon poselitev na naselbini vse od prazgodovinskih dob do srednjeveškega obdobja. Izkazalo se je, da se postojanka po velikosti, ohranjenosti arhitekture in kompleksnem značaju uvršča med najpomembnejše poznoantične naselbine v vzhodnoalpskem prostoru. Njena strateška lega in nadzor nad prehodom čez občutljivo območje tik pred vrati Italije pa sta opozorila tudi na verjeten vojaški značaj postojanke v poznorimskem obdobju in njegovo pomembno vlogo v sistemu poznoantične obrambe Italije. Zdelo se je, da je s tem naše delo na najdišču začasno končano. Vendar so zanimanje javnosti, velik interes kolegov in njihov odziv na prva poročila spodbudili tudi drugačna razmišljanja. Tako se je dilema, ali 7 Treba je korigirati tudi navedbe o začetkih zanimanja za najdišče, ki so bile objavljene leta 1995 v Varstvu spomenikov 35, 105. Konservatorka ZVNKD Nova Gorica N. Osmuk je v njem zapisala, da so nekateri lastniki zemljišč iz ogorčenja zaradi odnesenih predmetov, ki so jih našli sodelavci Inštituta za arheologijo ZRC SAZU, poskusno izkopali manjšo sondo. Kronološki potek dogodkov pa je bil v resnici povsem drugačen. Izkop na platoju s cerkvami (kot se je izkazalo kasneje) so domačini opravili že ob koncu osemdesetih let in v Goriški muzej odnesli primerke rimskih opek z najdišča, a pri arheologih niso zbudili želenega zanimanja. V svoji vnemi so zato izkopali jamo v velikosti 2 x 2 m, da bi pridobili več podatkov o najdišču. Prav to jamo, ki je bila že povsem zaraščena in prekrita z debelo plastjo listja in vejevja, smo leta 1993 očistili in dokumentirali. Ž. Šmit, ki je z iskalcem kovin odkril prve značilnejše kovinske najdbe, ni bil sodelavec Inštituta za arheologijo, kot se je konservatorki zapisalo, ampak je povsem neodvisno, brez vednosti članov Inštituta za arheologijo, opravil pregled takrat še časovno neopredeljenega najdišča, za katerega je izvedel prav iz notice N. Osmuk v Varstvu spomenikov. important position, and large Early christian complex with a number of churches clearly showing in the relief on the upper plateau of the settlement. A somewhat sensational description ('The Pompeii of Slovenia') that was published in Primorske novice by Zoran Božič, an enthusiastic visitor of the excavations at Tonovcov grad and archaeology aficionado, stirred great interest in the site, as well as provoked a response from the Kobarid dean Franc Rupnik (Božič 1993; Rupnik 1993). Even though the research was in a preliminary phase, a response from an archaeologist had to follow, for it was necessary to prevent the 'new Kobarid miracle' and the 'Slovenian Pompeii' from getting needlessly embedded into the minds of the Slovenian Antiquity enthusiasts. Summing up the existing research results Slavko ciglenečki explained the basic characteristics of the newly discovered Late Antique post and thus replied to the dilemmas that have arisen in the newspapers as regards the character of the site (ciglenečki 1993). In 1993 we performed the first research intervention in the settlement. We focused on an earlier unauthorised trench on the upper plateau of the settlement (between the central and south church) and cleared the severely collapsed edges of the pit which gave us an insight into the cultural layers.7 Apart from ruins of masonry structures we excavated a large number of Late Antique pottery fragments (coarse kitchenware, amphorae and African Red Slip Ware) and Antique rooftiles (imbreces and tegulae). The first more detailed results of the preliminary research were published in 1993 in the journal Kronika, a thematic issue of which was dedicated to the history of the Tolmin area (ciglenečki 1994b). Somewhat later 7 We should also correct the statement regarding the beginning of the interest for the site that was published in 1995 in Varstvo spomenikov 35, 105. N. Osmuk, the conservator from ZVNKD Nova Gorica wrote that some land owners were so appalled by the fact that the colleagues from the Institute of Archaeology at ZRc SAZU took away the objects they found, that they decided to dig a small trial trench. The chronological order of the events was quite different. The excavation on the plateau with the churches (as it was proven later) was performed by the locals at the end of the 1980s and they took samples of Roman bricks from the site to the Goriški muzej, however these did not raise the desired interest amongst the archaeologists. In their enthusiasm the locals dug a larger hole measuring 2 x 2 m, with the desire to obtain additional data on the site. It was this overgrown hole covered by a thick layer of leaves and branches that we cleared and documented in 1993. Ž. Šmit who discovered the first characteristic metal finds with a metal detector did not work for the Institute of Archaeology (as stated by the conservator), but independently, without the knowledge of the members of the Institute of Archaeology. He performed an overview of the site that was at the time still uncategorised (as regards its age) and about which he learned from the very article written by N. Osmuk and published in Varstvo spomenikov. sondažno izkopavanje spremeniti v sistematično, pokazala že kmalu po opravljenih začetnih delih. v Sloveniji je bilo že od sedemdesetih let 20. stoletja dalje odkritih nekaj poznoantičnih višinskih naselbin, od katerih so bile posamezne zaščitno, druge pa tudi sistematično raziskane. vendar pa se je pri teh kmalu izkazalo, da izkopavanjem večinoma niso sledile v enaki meri tudi sistematične in strokovno ovrednotene objave najdišč. Zato smo se seveda spraševali o smiselnosti odpiranja novega poznoantičnega najdišča. Inštitutska ekipa je prav v tem času raziskala in objavila dve pomembni arheološki poznoantični najdišči Kučar in Tinje (J. Dular, Ciglenečki, A. Dular 1995; Ciglenečki 2000), ki sta bili izkopani zaradi zaščitnih razlogov, tako da smo bili tako kadrovsko kot tudi strokovno dobro pripravljeni na sistematične raziskave. Te so pomemben segment dejavnosti Inštituta za arheologijo, ki naj razrešujejo posebej pomembne arheološko-historične probleme z usmerjenimi ciljnimi raziskavami. Iz dotedanjih izkušenj pa smo tudi vedeli, da raziskana najdišča z izjemo Rifnika in delno Ajdne niso bila tako močno vpeta v turistične tokove, da bi v slovenski zavesti spremenila pogled na izjemno pomembno obdobje transformacije antičnega sveta in na obstoj staroselcev. Temu botruje predvsem dejstvo, da leži večina poznoantičnih višinskih naselbin na odmaknjenih območjih, težko dostopnih krajih in zunaj sodobnih turističnih poti. Tonovcov grad pa vse te predpogoje v celoti izpolnjuje (lep, neokrnjen naraven ambient v najlepšem delu soške doline, turistična razvitost zgornjega Posočja, bližina Kobarida in magistralne ceste, bližina turistične poti, izredno aktivno Turistično društvo Kobarid itd.). Vse našteto, kot tudi izdatna pomoč članov Muzejskega društva in Občine Kobarid, so napeljevali na misel, da bi bilo na Tonovcovem gradu mogoče predstaviti pomembno in z arhitekturo zelo bogato poznoantično naselbino, ki bi jo bilo mogoče v celoti vključiti v turistično ponudbo soške regije in s tem njene ostaline ter problematiko, ki jo ponazarja, predstaviti širšemu krogu prebivalstva, ne samo arheologom in zainteresiranim laikom. Tehtanje za in proti je na koncu odločilo v prid sistematični raziskavi najdišča, hkrati pa smo se začeli pogovarjati tudi z vsemi pristojnimi ustanovami o njegovi konservaciji in učinkoviti predstavitvi izsledkov naših raziskav. Skratka, najdišče se je zaradi vsega navedenega zdelo najprimernejše za sistematično raziskavo, hkrati pa je ponujalo možnost prezentacije v samem središču turističnih tokov, s čimer bi lahko prispevalo k uveljavitvi tovrstnih najdišč in posredno popularizaciji prej slabo znanega obdobja zatona antike ter njegove kompleksne etnične situacije. Zato smo pripravili predlog načrta sistematične raziskave in ga predložili Ministrstvu za znanost RS, ki nam je odobrilo denar za začetne raziskave. Finančno in we published a slightly edited version of the translated text in Italian (Ciglenečki 1994c). The preliminary research thus roughly defined the time span of the settlement from Prehistory to the Middle Ages. We realized that with its size, state of preservation and its complex character the post belongs amongst the most important Late Antique settlements in the East Alpine area. Its strategic position that controls the pass through a vulnerable area at the very gateway to Italy pointed to the possible military character of the post during the Late Roman period and its important role in the defence of Italy. It seemed that our work at the site temporarily ended. However, the interest of the general public, the great interest of our colleagues and their response to the first reports encouraged a different line of thought. Thus soon after the initial research we faced the dilemma whether to change the trial trenching into systematic excavations. From the 1970s on quite a number of Late Antique hilltop settlements were discovered in Slovenia. On some of them protective excavations were carried out, while others were systematically researched. However, with these it was soon noticed that the excavations were not necessarily followed by systematic and expertly evaluated publications. This of course made us consider whether it was at all sensible to open a new Late Antique site. At that time the team from the Institute had just finished researching and publishing two important archaeological sites (Kučar and Tinje, both of which were excavated due to protective reasons; J. Dular, Ciglenečki, A. Dular 1995; Ciglenečki 2000), thus we had experts and knowledge to start systematic research. They represents an important segment of the activities carried out by the Institute of Archaeology, which should solve the particularily important archeo-historical problems through targeted research. We were aware that the previously researched sites (with the exception of Rifnik and partially Ajdna) were not enough incorporated into the tourist routes to change the general views on the extremely important period of the transformation of the Antique world and the existence of the autochthonous population. These views are the consequence of the fact that most Late Antique hilltop settlements are situated in remote areas, hard to access places and outside of the tourist paths. Tonovcov grad, on the contrary, fulfils all the conditions for a popular tourist spot (unspoilt landscape in the most beautiful part of the Soča Valley, already popular with tourists, proximity to Kobarid, the main road and the tourist paths, is under the jurisdiction of an exceptionally active Tourist association of Kobarid, etc.). All of this, along with the substantial help provided by the members of the Museum Society and the Municipality of Kobarid, lead to the idea that the important and architecturally extremely rich settlement on Tonovcev grad could be properly presented and included into the tourist offer preko sistema javnih del sta nam pri prvih sistematičnih raziskavah pomagali tudi Turistično društvo Kobarid (predsedniki Pavel Sivec, Vojko Hobič, Željko Cimprič) in Upravna enota Tolmin (načelnik Zdravko Likar), pozneje pa še občina Kobarid (župana Pavel Gregorčič in Robert Kavčič), Fundacija Poti miru v Posočju (predsednik uprave Zdravko Likar) in predvsem Kobariški muzej (direktor Jože Šerbec). V letu 1994 je bilo sistematično izkopano območje velike stanovanjske stavbe, ki je bila edina že poškodovana, da bi se tako kar najbolje pripravili na nadaljnja raziskovanja in dodobra spoznali stratigrafijo najdišča. Tega leta je bila ustanovljena tudi posebna konservatorska komisija, ki je bedela nad izvajanjem konservatorskih del in pripravila načrt prezentacije najdišča.8 Prvič se je sestala 2. junija 1994 ter uspešno spremljala dela in sproti pripravljala smernice za konservacijo in prezentacijo. Restavratorski center iz Ljubljane (odgovorna konservatorja Franc Vardijan (t 2006) in Jernej Hudolin) je leta 1996 v celoti prevzel vsa konservatorsko-restavratorska dela na tem arheološkem spomeniku. Leta 1995 je bil izdelan dolgoročni program sistematičnih raziskav in konservatorskih posegov na najdišču, ki je okvirno določal obseg in tempo raziskovanja. V tem letu je bil izkopan, v skladu z že prej zastavljeno mrežo kvadrantov, vmesni del med že raziskano stanovanjsko stavbo ter platojem s cerkvami. Sakralni sklop kot najzahtevnejši del dosedanjih raziskav je bil izkopan v letih 1996 in 1997, vmesni del med osrednjo in južno cerkvijo pa v letih 2003 in 2004. Leta 1999 smo skušali locirati grobišče, zato smo naredili tri sonde na terasi na jugovzhodnem pobočju Tonovcovega gradu, ob nekdanjem Tonovcovem seniku. Vse sonde so bile negativne. V letih 2002, 2003 in 2006 je bil izkopan velik zbiralnik za vodo, ki je bil poškodovan že med drugo svetovno vojno, ko je bil v njem zgrajen bunker in je vanj vodil strelski rov. Raziskave cerkvenega kompleksa smo končali v letih 2002 in 2005, ko so bili raziskani ostanki arhitekture tik pod severno cerkvijo, ki jih je bilo zaradi bližine cerkva hipotetično mogoče povezati z njimi. Objekta sta bila označena s številkama 2 in 3 (sl. 1.7) in sta tako zaokrožila podobo poselitve na skalnem platoju in tik pod njim. Raziskovanja so z manjšim ali večjim obsegom tako tekla pravzaprav kontinuirano od leta 1993 dalje in so bila prekinjena le s sezonami, v katerih je bilo treba zidane stavbe konsolidirati in - v primeru severne in osrednje cerkve - zavarovati z zaščitno zgradbo (arhitektka Veronika Ščetinin; sl. 1.13, 1.14) of the Soča region. Thus the remains and the issues they represent would be available to the general public and not only to the archaeologists and enthusiasts. In the end we decided in favour of a systematic excavation, and at the same time we started a dialogue with relevant institutions regarding the preservation of the site and an a good visual presentation of our research findings. Due to all this the site seemed most appropriate for a systematic research and at the same time it offered the possibility of presentation in the very heart of tourist routes, which could help popularise such sites and consequently the previously poorly known period of the decline of Antiquity and its complex ethnical situation. So we prepared a proposal for systematic investigation and presented it to the Ministry of Science of the Republic of Slovenia which funded the preliminary research. In the first systematic research we were financially (as well as through the public works system) helped by the Kobarid Tourist Association (presidents Pavel Sivec, Vojko Hobič, Željko Cimprič) and the Tolmin Administrative Unit (chief Zdravko likar), and later on also by the municipality of Kobarid (the mayors Pavel Gregorčič and Robert Kavčič), foundation Poti miru v Posočju (president of the board Zdravko Likar) and especially the Kobariški muzej (director Jože Šerbec). In 1994 we systematically excavated the area of the large building which had been previously damaged and thus seemed the best place to get acquainted with the stratigraphy of the terrain and prepare for future research. A special conservation committee was also established in the same year to supervise the conservation works and prepare the plan for the presentation of the site.8 It assembled for the first time on 2nd June 1994 and successfully monitored the work as well as prepared the guidelines for the preservation and presentation. In 1996 the Restoration Centre from Ljubljana (the responsible conservators Franc Vardijan (t 2006) and Jernej Hudolin) took over all preservation and restoration works on this archaeological monument. In 1995 the long term plan for the systematic excavations and preservation interventions was elaborated which roughly defined the scope and tempo of the excavations. In that year the area between the previously invstigated building and the plateau with the churches was excavated according to the previously planned grid. The ecclesiastical complex, which represented the most demanding part of the research so far, was excavated in 1996 and 1997, while the part between the central and south church was explored in 2003 and 2004. 8 člani komisije so bili v začetku N. Osmuk iz ZVNKD Nova Gorica kot pristojna konservatorka, M. Slabe, F. Vardijan, 1. Bogovčič, M. Sagadin, 1. Curk, D. Vuga in S. Ciglenečki, pozneje pa še J. Hudolin in P. Bratina. 8 At the beginning the members of the committee were the following: N. Osmuk from ZVNKD Nova Gorica as the conservator in charge, M. Slabe, f. Vardijan, 1. Bogovčič, M. Sagadin, I. Curk, D. Vuga and S. Ciglenečki, and later on also J. Hudolin and P. Bratina. Sl. 1.13: Tonovcov grad. Pogled z vzhoda na zaščitni zgradbi nad severno in osrednjo cerkvijo. Fig. 1.13: Tonovcov grad. A view of the protective constructions above the north and central church, from the east. Že med potekom večjih raziskav smo najdišče skupaj s sodelavci Kobariškega muzeja (predvsem J. Šerbec) opremili s kažipoti, ki so usmerjali obiskovalce, in pripravili informacijske panoje. Leta 2007 sta bili v celoti dokončani zaščitni zgradbi nad bolje ohranjenima cerkvenima stavbama, urejeni so bili zaščitni hodniki, ki omogočajo ogled notranjosti, leta 2008 pa postavljena še zaščitna streha nad vodnim zbiralnikom. Sakralne zgradbe in ostala že izkopana poslopja so bila zavarovana in urejena, tako da omogočajo obiskovalcem prvi uvid v arhitekturo in ambient, v Tolminskem in v Kobariškem muzeju pa so predstavljeni tudi nekateri izbrani predmeti, ki dovolj nazorno ilustrirajo življenje v poznoantični naselbini. Dosedanji obisk, ki se iz leta v leto veča, kaže številke, kakršne druga arheološka najdišča na našem območju redko dosegajo (pribl. 20.000 obiskovalcev na leto), in dokazuje, da je bila odločitev za raziskavo Tonovcovega gradu pravilna ter da v celoti izpolnjuje v izhodišču zadane cilje. In 1999 we tried to locate the cemetery, thus three trial trenches were made along the terrace on the southeast slope of Tonovcov grad, alongside the former hayloft. All trial trenches were negative. In 2002, 2003 and 2006 a large water reservoir was investigated. It had been damaged during World War II, when a bunker and a military trench were constructed within the preserved structure. The research of the ecclesiastical complex was concluded in 2005, when the architectural remains under the plateau to the north of the north church were excavated. Due to the proximity of the churches these remains could at least hypothetically be linked to the churches. The buildings were given numbers 2 and 3 (Fig. 1.7) and rounded up the image of the structures on the rock plateau and immediately below it. Thus the research continued (more or less intensively) since 1993 with the exceptions of the seasons in which the masonry buildings needed to be consolidated and - in the case of the north and central church - protected with a protective construction (architect Veronika Ščetinin; Figs. 1.13, 1.14). The site was equipped with signposts and informational panels for visitors (all of which were prepared together with our co-workers from the Kobariški muzej in Kobarid, especially J. Šerbec) during the excavations. Sl. 1.14: Tonovcov grad. Odlično ohranjena in z moderno zaščitno zgradbo zavarovana notranjost severne cerkve. Fig. 1.14: Tonovcov grad. The excellently preserved interior of the north church protected by the modern construction. In 2007 the protective buildings that covered the better preserved churches were completed, protective hallways that give the view of the interior were created, and in 2008 a protective roof was raised above the water reservoir. The churches and the remaining excavated buildings were protected and presented to offer the visitor an insight into the architecture and the interior. In the Tolminski muzej in Tolmin and in the Kobariski muzej in Kobarid some artefacts were put on display in order to illustrate the life in the Late Antique settlement. As the number of visits has been growing each year, and the site has been achieving numbers that other archaeological sites in Slovenia find hard to reach (approximately 20.000 visitors per annum), we are convinced that the decision to explore Tonovcov grad was correct and that it fulfils all the goals set at the beginning. 1.5 posočje v poznoantičnem času 1.5 the posočje area during late antiquity Raziskave poznorimskega zapornega sistema Claustra Alpium luliarum so za dolgo obdobje določale raziskovalne prioritete slovenskih in tudi nekaterih italijanskih in avstrijskih arheologov. Ostala arheološka dejavnost v zahodni Sloveniji je bila tako dolgo v senci enega najpomembnejših arheoloških spomenikov v alpskem območju. Raziskovati so ga začeli že v drugi polovici 19. st., vrh pa so raziskave doživele s sistematičnimi izkopavanji v 60. in 70. letih 20. st. (pregled pri Šašel, Petru 1971). Z raziskovanjem utrdb zapornega sistema se je povečalo tudi zanimanje za ostale poznoantične postojanke v njihovem zaledju kot tudi pred njim. Zastavljene so bile raziskave na nekaj ključnih točkah, ki omogočajo razumevanje preživetja staroselcev v nemirnem času pozne antike in preseljevanja ljudstev (Petru, Ulbert 1975; Bolta 1981). Ta raziskovalna usmerjenost se je Posočja dotaknila le bežno, saj je bila večina raziskav v zahodni Sloveniji osredotočena vzdolž glavne rimske vpadnice iz Emone v Akvilejo in na območju Gorenjske. Prve skromne poskuse sistematičnega arheološkega raziskovanja poznoantičnega obdobja v Posočju z namenom razumeti poselitvene razmere, je mogoče zaznati šele s sondiranjem na Sv. Katarini (Kekec) pri Novi Gorici v letih 1971 in 1976, vzporedno z velikimi izkopavanji, ki so takrat potekala na Hrušici (Svoljšak 1990). Nenavaden kultni prostor na pobočju Gradiča nad Kobaridom, ki so ga uporabljali vse do začetka 5. st. in odkritje izjemno dobro ohranjene utrjene po-znoantične naselbine na Tonovcovem gradu so žarišče antičnih raziskav v zadnjih dveh desetletjih postavili prav v to, prej tako slabo raziskano območje (Osmuk 1987a; Ciglenečki 1994b). Poglejmo, kakšne rezultate nam ob izhodiščni navezavi na Tonovcov grad kot referenčno točko par excellence ponuja Posočje z množico skromnih podatkov, ki jih je mogoče bolje razumeti prav v luči raziskav tu obravnavanega najdišča. Naselitvena podoba poznoantičnega obdobja v Posočju je dokaj pestra in razkriva podobno kot v večjem delu slovenskega prostora široko paleto naselbinskih tipov, vse od še obstoječih nezavarovanih naselij v 4. st. pa do močno utrjenih bivalnih središč na naravno zavarovanih položajih v 6. st. (sl. 1.15). Ob tem je treba že For a long period the research priorities of Slovenian as well as certain Italian and Austrian archaeologists were focused on the Late Roman defence system known as Claustra Alpium luliarum. The remaining archaeological activities in western Slovenia thus remained in the shade of one of the most important archaeological monuments in the Alpine area. This system was researched as early as the second half of the 19th century, however the research peaked with the systematic excavations carried out in the 1960s and 1970s (see Šašel, Petru 1971). With the research of the fortifications found within the defence system the interest in other Late Antique posts in the system's hinterland as well as in front of it started growing. Research was initiated into the key strongholds that enable the understanding of the survival of the autochthonous population during an extremely factious period of the Late Antiquity and the Migration Period (Petru, Ulbert 1975; Bolta 1981). This research only briefly dealt with the Posočje area, for most of the research in western Slovenia was focused on the area around the main Roman road leading from Emona towards Aquileia, and the Gorenjska region. The first modest attempts of systematic archaeological research of the Late Antiquity in the Posočje area - with the intent of understanding the settlement conditions - took place as late as 1971 and 1976 when trenching was carried out on Sv. Katarina (Kekec) above Nova Gorica at the same time as extensive excavations took place on Hrušica (Svoljšak 1990). The unusual cult place on the slope of Gradič above Kobarid that was used until the beginning of the 5th century and the extremely well preserved fortified Late Antique settlement at Tonovcov grad set the focus of the Antiquity research during the last two decades in this previously poorly researched area (Osmuk 1987a; Ciglenečki 1994b). Let's take a look at what sort of results the Posočje area with its links to Tonovcov grad as a referential point par excellence offers us with its abundant yet modest data that can be better understood through the research of the here discussed site. The settlement image of the Late Antiquity in the Posočje region is relatively diverse and - similar to many parts in Slovenia - reveals a wide palette of settlement Sl. 1.15: zemljevid Posočja s poznoantičnimi arheološkimi najdišči. Fig. 1.15: Map of the Posočje area with Late Antique sites. uvodoma poudariti, da je ta podoba znana predvsem z nekaj bolje raziskanih območij (Kobarida, Mosta na Soči in Nove Gorice), medtem ko imamo od drugod na razpolago skromne in večinoma naključne podatke, ki ne omogočajo celovitega razumevanja tedanje poselitve. v vzhodnoalpskem območju prevladuje - gledano z vidika poselitve - okvirna omejitev poznoantičnega obdobja z letnicama 300-600 (Demandt 1989, XIX-XXI; Ciglenečki 1999, 289). Novejše raziskave pa so pokazale, da je do prvih pomembnejših naselbinskih premikov prišlo že v zadnji tretjini 3. st., in sicer zaradi vedno močnejšega pritiska barbarov na meje imperija. To je odsevalo v kratkotrajni obiskanosti starejših prazgodovinskih gradišč, gradnji številnih mestnih obzidij, hkrati pa je botrovalo nastanku novih vojaških utrdb (Ciglenečki 1990). V Posočju smo zanesljivo sled teh prvih premikov na višinske položaje ugotovili le na Tonovcovem gradu. Zaradi pretežno hribovitega sveta je bilo v Posočju ugotovljenih le nekaj vil rustik, ki bi jim bilo mogoče dokazati obstoj ali celo nastanek v poznorimskem času. Marsikje podatki sicer kažejo na njihov obstoj, raziskane so bile redke. Med bolj znanimi omenimo le rimski vili v Bukovici in Pavlinih pri Loki v zahodnem delu Vipavske doline (Osmuk 1981; Žbona Trkman 1987). Bolje raziskana je bila le slednja, ki pa je bila opuščena ali uničena prav v drugi polovici 3. st., torej že v času prvih usodnejših naselbinskih premikov. zidove, ki jih je mogoče opredeliti kot stanovanjsko zgradbo iz 4. st., so odkrili na Ledinah v Novi Gorici (Osmuk 1987b). več je arheoloških sledov, ki posredno nakazujejo obstoj poznorimskih nezavarovanih naselbin. Takšni sta predvsem manjša nekropola iz 4. st. Na lajšču v Avčah in podobna v Anhovem (ANSl 1975, 124; Kos 1988, 66, št. 27). Med zaščitnim izkopavanjem nekropole Na steni v Solkanu so odkrili nekaj poznorimskih grobnih celot, ki pa še niso objavljene (Knific, Svoljšak 1984, 277). Glede na velikost in časovno opredelitev omenjenih grobov je mogoče sklepati na skromne pripadajoče neutrjene naselbine iz 4. st. Obstoj bivališč ob Idrijci potrjujeta posamezna poznorimska grobova iz sicer večjih nekropol, katerih začetki segajo že v latensko obdobje: grob iz Idrije ob Bači in iz Slapa ob Idrijci (Kos 1988, 23). Skratka, podoba starejših, nezavarovanih naselbin iz 4. st. je povsem neznana, posredno potrjena le z nekaj grobišči, ki nakazujejo njihovo prisotnost. Arheološke indikacije potrjujejo obstoj dveh pomembnih naselbin, pri katerih pa ni mogoče povsem zanesljivo opredeliti značaja in obsega poselitve: prva je Gradič nad Kobaridom in druga Most na Soči. Večletna raziskovanja na Gradiču nad Kobaridom že omogočajo okvirno kronološko zamejitev trajanja poselitve. Rimska naselbina na pobočjih in temenu Gradiča je nastala v območju nekdanjega velikega prazgodovinskega gradišča in se nadaljevala - verjetno celo brez večje časovne cezure - v 4. st. in delno še v začetek 5. st. (s1. types ranging from the still existing unprotected settlements from the 4th century to the strongly fortified settlements on naturally protected positions dated to the 6th century (Fig. 1.15). Of course, it should be emphasised that this image is known from the slightly better researched areas (Kobarid, Most na Soči and Nova Gorica), while elsewhere only modest and mainly accidental data is at our disposal, and this does not allow for a wholesome understanding of the settlement of the time. As regards settlement in the Eastern Alpine area the Late Antiquity is most commonly marked by the period between 300 and 600 AD (Demandt 1989, XIX-XXI; Ciglenečki 1999, 289). Newer research has shown that the first important settlement changes occurred already during the last third of the 3rd century. These changes occurred due to the increasing pressure of the Barbarians upon the Empire borders. This was reflected in the short visits of the older prehistoric settlements, the construction of numerous town walls, as well as the appearance of new military fortifications (Ciglenečki 1990). In the Posočje area only Tonovcov grad represented a reliable proof of these first moves towards the highland positions. Due to the predominantly mountainous world of the Posočje area only a few Roman villas were discovered that existed or were built during the Late Roman period. Even though there is data that indicates the existence of villas only a handful of them have been researched so far. Amongst the best known are the Roman villas in Bukovica and Pavlini near Loka in the western part of the Vipava Valley (Osmuk 1981; Žbona Trkman 1987). Only the latter one, which was abandoned or destroyed in the second half of the 3rd century (i.e. during the period of the first important moves), has been researched in greater detail. Walls that can be defined as belonging to 4th century living quarters were discovered at Ledine in Nova Gorica (Osmuk 1987b). There are a number of archaeological indicators that indicate the existence of unprotected Late Roman settlements. Examples of such are the small 4th century necropolis at Na lajšču in Avče and a similar one in Anhovo (ANSl 1975, 124; Kos 1988, 66, No. 27). A few Late Roman graves were discovered during the protective trenching at the necropolis Na steni in Solkan, however these finds remain unpublished (Knific, Svoljšak 1984, 277). Taking into account the size and dating of the aforementioned graves it can be concluded that they were accompanied by modest unfortified settlements dating to the 4th century. The existence of a settlement close to the Idrijca river was confirmed by two individual Late Roman graves from two otherwise larger cemeteries, the beginnings of which reach into the La Tene period: i.e. the graves from Idrija pri Bači and from Slap ob Idrijci (Kos 1988, 23). Thus the layout of the older and unprotected 4th century settlements is entirely unknown and only circumstantially confirmed by a few cemeteries that indicate their presence. Sl. 1.16: Gradič nad Kobaridom. Pogled na gradišče s severovzhodne strani. Fig. 1.16: Gradič above Kobarid. A view of the hillfort from the northeast. Sl. 1.17: Gradič nad Kobaridom. Pogled na gradišče z južne strani. V ozadju Tonovcov grad. Fig. 1.17: Gradič above Kobarid. A view of the hillfort from the south. Tonovcov grad in the background. 1.16-1.18). Poznoantične zgradbe doslej sicer niso bile raziskane, vendar množica novcev in drugih drobnih najdb v območju domnevnega svetišča posredno priča o njihovem obstoju (Osmuk 1987a; 1997a). Tukajšnji izsledki dokazujejo, da je naselbina obstajala (morda v zmanjšanem obsegu?) še v času, ko na Tonovcovem gradu ugotavljamo prisotnost močnejše vojaške posadke in trajnejše sledove bivanja. v Mostu na Soči so ob izkopavanjih velike in izjemno ohranjene prazgodovinske naselbine naleteli Archaeological traces confirm the existence of two important settlements, however their character and scope cannot be precisely defined: the first is Gradič above Kobarid and the other is Most na Soči. The long lasting research at Gradič above Kobarid allows for a partial chronological determination of the duration of the settlement. The Roman settlement on the slopes and ridge of Gradič emerged in the area of a former large prehistoric site and continued - most likely without a longer time gap - into the 4th and partially into the be- Sl. 1.18: Gradič nad Kobaridom. Poskus zarisa obsega gradišča po lidarskem posnetku (© ZRC SAZU). Fig. 1.18: Gradič above Kobarid. Mapping the hillfort on a lidar data visualization (© ZRC SAZU). tudi na številne rimske sledove (sl. 1.19). Iz preliminarnih podatkov je mogoče izluščiti omembo osmih rimskih stavb, bolje znana je le ena. V obravnavanem kontekstu je pomembno, da so ob številnih drugih sledovih odkrili več rimskih novcev in tri poznorimske zakladne najdbe. Prvi dve, ki sodita v drugo polovico 4. st., sta bili najdeni v ruševinski plasti rimske stavbe, izkopane leta 1979, tretja, katere zakop je natančno datiran v november leta 401, pa je bila odkrita na ognjišču rimske stavbe (Kos 1988, 31-32, št. 9/3 in 9/4; 33, št. 9/5). Njihov zakop odlično nakazuje trajanje rimske naselbine v začetek 5. st., kar je omogočila naravno odlično zavarovana lega na sotočju Soče in Idrijce, v celoti pa tej dataciji pritrjuje grobišče s 149 grobovi, od 1. do 5. st. (Horvat 1999, 253). Pri obeh zadnjih obravnavanih naselbinah ni mogoče zanesljivo ugotoviti, ali gre za nezavarovani naselbini ali pa so prebivalci vsaj deloma že izkoristili njuno ugodno obrambno lego: na Gradiču delno ginning of the 5th century (Figs. 1.16-1.18). So far Late Antique buildings have not been researched, however the high numbers of coins and other small finds in the area covered by an assumed cult place indicates their existence (Osmuk 1987a; 1997a). The data from this site indicates that a settlement still existed (maybe on a smaller scope?) at a time when a stronger garrison and traces of more permanent settlement were discovered at Tonovcev grad. In Most na Soči the excavations of the large and extremely well preserved prehistoric settlement also revealed numerous Roman traces (Fig. 1.19). The preliminary data indicates the existence of eight Roman buildings, however only one of them has been researched into any sort of detail. In this context it is important to mention that a number of Roman coins and three Late Roman hoards were discovered amongst numerous other traces. The first two hoards can be dated to the second half of the 4th century and were discovered in Sl. 1.19: Most na soči. Načrt najdišča z rdeče označenimi rimskimi stavbami (po Maggi, Žbona Trkman 2007, sl. 7). Fig. 1.19: Most na soči. plan of the site with Roman structures marked in red (after Maggi, Žbona Trkman 2007, Fig. 7). znotraj starih prazgodovinskih okopov in na Mostu na Soči v prostoru, odlično zavarovanem z rekama, ki ga je bilo mogoče na najlažje dostopnem delu tudi uspešno braniti. Veliko bolje je razvidna poselitev na utrjenih višinskih položajih, ki se je polno uveljavila v tem času (Ciglenečki 1999, 291-294). Prepoznavanje teh naselbin je olajšano z lego na že naravno dobro zavarovanih krajih, hkrati pa so bile zaradi svoje po navadi odmaknjene hribovske lege pozneje manj poškodovane kot sočasne ravninske naselbine. Sistematične raziskave na Tonovcovem gradu so dokazale, da je bila tam v drugi polovici 4. st. nameščena močna vojaška posadka, katere naloga je bila predvsem kontrola prehoda proti Italiji in varstvo rimske ceste (glej pogl. 5.1). Podobne postojanke so takrat nastale tudi na nekaterih drugih naravno zavarovanih hribih, kjer pa manjkajo zanesljivi elementi za podrobnejšo označitev njihove funkcije. Vse označujejo naravno zavarovana lega, ponekod odkrite poznorimske ali tudi že starejše obrambne naprave, hkrati pa arheološki predmeti, ki dovoljujejo podrobnejšo datacijo. takšno je pred nedavnim odkrito najdišče Selce pri Zatolminu (sl. 1.20), ozek gozdnat pomol, ki se strmo dviga nad ravnico the destruction layer of the Roman building excavated in 1979, while the third, for which it has been successfully ascertained that it was buried in November 401, was discovered on the fireplace of a Roman building (Kos 1988, 31-32, Nos. 9/3 and 9/4; 33, No. 9/5). Their burial indicates that the Roman settlement continued into the beginning of the 5th century, which was enabled by the naturally excellently protected position at the confluence of the soča and Idrijca rivers. This has also been confirmed by the graveyard with 149 graves that can be dated between the 1st and 5th centuries (Horvat 1999, 253). At both of the settlements mentioned above it is impossible to precisely determine whether they were unsecured settlements or whether the settlers made at least partial use of their favourable defensive position: on Gradič partially with the old prehistoric ramparts and on Most na soči by the two rivers that offer an excellent protection of the location that could be successfully defended at the point with the easiest access. Much clearer is the settlement pattern of fortified hilltops that was already fully established at the time (Ciglenečki 1999, 291-294). Recognising these settlements is made easier by their position on naturally well Sl. 1.20: selce pri zatolminu. Pogled na poznoantično postojanko z južne strani Fig. 1.20: selce near zatolmin. A view of the Late Antique post from the south. Sl. 1.21: sv. Katarina nad Novo Gorico. Pogled na poznoantično utrdbo z južne strani. Fig. 1.21: sv. Katarina above Nova Gorica. A view of the Late Antique fortification from the south. (Cvitkovič 1999, 56; Mlinar, Knavs 2004). Nekaj kovinskih najdb (predvsem fibula s čebulastimi gumbi) datira najdišče v poznorimsko obdobje. Ali gre za strateško nameščeno opazovalnico, pribežališče v dolini prebivajočih staroselcev ali celo kombinacijo obojega? Poznorimska naseljenost in pripadnost obrambnih naprav sta vprašljivi tudi na velikem kompleksnem najdišču Sv. Katarina nad Novo Gorico. velik kopast hrib, na katerem je nekdaj stala cerkev sv. Katarine, z n. m. v. 328 m, je prislonjen na južno pobočje Škabri- protected locations and due to the fact that their usually remote highland position did not make them as prone to being destroyed later on as the settlements in the lowland from the same period. The systematic research at Tonovcov grad has shown that in the second half of the 4th century this was the home to a garrison, the task of which was to control the access route towards Italy and protect the Roman road (see chapter 5.1). At the time similar posts emerged on other naturally protected hills, however in Sl. 1.22: Sv. Katarina nad Novo Gorico. Načrt naselbine (Svoljšak 1990, sl. 3). M. = 1:5000. Fig. 1.22: Sv. Katarina above Nova Gorica. Settlement plan (Svoljšak 1990, Fig. 3). Scale = 1:5000. jela in strateško zelo ugodno umeščen pri izteku Soče s hribovja v ravnico (sl. 1.21). Na treh straneh je odlično zavarovan s strmimi pobočji in deloma s skalami, le na severni strani ga nekaj metrov nižje sedlo loči od višjega hribovja za njim. Približna velikost zavarovanega prostora znaša 250 x 160 m (sl. 1.22). Veliko prazgodovinsko gradišče je bilo v poznoantičnem času ponovno utrjeno. starejši avtorji omenjajo poznoantično obzidje in kar štiri dobro vidne obrambne stolpe. Izkopavalec Drago Svoljšak je s sondiranji leta 1971 in 1976 ugotovil dve fazi poznejših gradenj na prazgodovinskem okopu, 1,4 m debelo obzidje in obrambni stolp, katerega velikost ni znana (Svoljšak 1990; Osmuk 1997b, 176). Najdba zlizanega novca iz 4. st. ne more biti dovolj tehten razlog za datacijo prve faze obzidja v poznorimski čas, a bi smeli glede na sočasne primerjave z vzhodnoalpskega območja v starejši fazi vendarle pomisliti na obzidje iz 4. ali začetka 5. st., kot se je to izkazalo drugod (Rifnik, Duel, Zbelovska gora, Šumenje itd.). To bi potrjevala tudi debelina zidu, ki je bila tako močna le pri starejših poznorimskih utrdbah (Gradišče v Vrhniki, Velike Malence, Hrušica), medtem ko pri poznejših postojankah ne presega širine 0,80 m (Ciglenečki 1987a, 129). časovna pripadnost izstopajočih obrambnih stolpov je vprašljiva, njihovo izgradnjo bi smeli domnevati šele v zadnjem poznoantičnem obdobju (podobno Rifnik). the other cases no reliable elements that would enable us to precisely define their function were discovered. All are marked by the naturally protected position, and some have also revealed Late Roman or even older defence elements, as well as archaeological objects that lead to a precise dating. Such was the recently discovered site Selce near Zatolmin (Fig. 1.20), a narrow forest ridge that rises steeply above the plain (Cvitkovič 1999, 56; Mlinar, Knavs 2004). A few metal finds (especially the crossbow fibula) date the site into the Late Roman period. Was this a strategically placed observation point, a refuge of the autochthonous population living in the valley or a combination of both? Late Roman settlement and the defensive elements are questionable also on the large and complex site of Sv. Katarina above Nova Gorica. With an altitude of 328 metres above sea level, the large rounded hill that used to be the home to the church of St. Catherine and that leans upon the southern slope of Škabrijel is strategically excellently placed at a point where the river Soča runs from the mountains into the plains. Steep slopes and rocks on three sides offer excellent protection, and it is only on the northern side that the few metres lower saddle divides it from the higher hills behind it. The approximate size of the protected area measures 250 x 160 m (Fig. 1.22). The large prehistoric site was refortified in the Late Antiquity. Older authors mention Late Antique walls and four clearly visible defensive towers. With his trenching in 1971 and 1976 the excavator Drago Svoljšak ascertained two phases of later buildings on the prehistoric rampart: a 1.4 m thick wall and a defensive tower, the size of which is unknown (Svoljšak 1990; Osmuk 1997b, 176). The discovery of the well used 4th century coin cannot represent a sufficient reason to date the first phase of the wall into the Late Roman period, however keeping in mind the contemporary examples from the Eastern Alpine region we could consider this to be a 4th or early 5th century wall as this has proven to be the case elsewhere (Rifnik, Duel, Zbelovska gora, Šumenje, etc.). This is also confirmed by the thickness of the wall that was matched only at the older Late Roman fortifications (Gradišče in Vrhnika, Velike Malence, Hrušica), while it never surpassed 0.80 m at younger fortifications (Ciglenečki 1987a, 129). The chronological placement of the protruding defence towers is questionable and their construction should be assumed to belong into the last stages of the Late Antiquity (similarly Rifnik). The link between the Late Roman graveyard in Kosoveli and the settlement Tabor above Črniče 200 m away cannot be questioned. As indicated by individual Roman ceramics and especially a jug with burnished decoration the settlement was built on the location of a Late Roman fortification (Zavrtanik 1984; Knific 2004, 9). As a certain addition to the settlement pattern we have to take into account the prehistoric site Kozmac near Gojače that lies in the Vipava Valley and is situated Sl. 1.23: Kozmac. Načrt poškodovane poznorimske postojanke (po Harej 1988-1989, sl. 1). M. = 1:5000. Fig. 1.23: Kozmac. Site plan of the partly destroyed Late Roman post (after Harej 1988-1989, Fig. 1). Scale = 1:5000. Povsem zanesljiva je povezava med poznorimskim grobiščem v Kosovelih in 200 m oddaljeno naselbino Tabor nad Črničami, ki je bila zgrajena na mestu poznorimske utrdbe, kot nakazujejo posamezne najdbe rimske keramike in še posebej vrč z zglajenim okrasom (Zavrtanik 1984; Knific 2004, 9). Kot svojevrstno dopolnitev poselitvene slike je treba pritegniti tudi že v Vipavski dolini ležeče prazgodovinsko gradišče Kozmac pri Gojačah, situirano na povsem nizkem, a prostranem, danes delno uničenem griču (višina griča pribl. 20 m; sl. 1.23). Izsledki raziskav pričajo o skromni poznoantični poselitvi in dovoljujejo datacijo postojanke v čas 4. in tudi poznega 6. oz. celo 7. st. (Harej 1989). Njegova lega na nizki vzpetini povsem v bližini državne rimske ceste bi utegnila kazati na manjšo zasilno vojaško postojanko ali pribežališče, ki se je naslonilo na stare prazgodovinske okope. Zelo podobna se zdi problematika najdišča Sv. Marije na Jezeru pri Golem Brdu, kjer so na strateško umeščenem gradišču odkrili dva poznorimska novca: nakazujeta možnost, da je bil naravno zavarovan in v prazgodovini umetno utrjen hrib za kratek čas uporabljan tudi v drugi polovici 4. in na začetku 5. st. (Šemrov 2004, 35, št. 12; Bratina 1999). čeprav smo se v svojem pregledu omejili na slovenski del Posočja, pa je zaradi boljšega razumevanja tedanjih razmer na tem območju koristno pritegniti tudi pomembno bližnjo postojanko Gradina pri Doberdobu (Castelazzo di Doberdo), ki je bila le delno raziskana, vendar velike količine novcev z nje omogočajo zanesljivejši kronološki okvir (Furlani 1969; Maselli Scotti on a low, but vast, today partially ruined elevation (approximately 20 metres high; Fig. 1.23). The finds indicate a modest Late Antique settlement and allow for the post to be dated into the 4th and also into the late 6th or even 7th century (Harej 1989). Its position on a low elevation in the immediate vicinity of the Roman state road could indicate a small provisional military post or a refuge that was based on old prehistoric ramparts. The site Sv. Marija na Jezeru in the vicinity of Golo Brdo, where two Late Roman coins were discovered on a strategically placed prehistoric hillfort seems similar. The coins indicate the possibility that the naturally and in prehistory artificially fortified hill was for a short time used also in the second half of the 4th and the beginning of the 5th century (Šemrov 2004, 35, No. 12; Bratina 1999). Even though our overview is limited to the Slovenian side of the Posočje region we should mention the important nearby post Castelazzo near Doberdo, for this will shed additional light on the conditions of the time. This post has been partially researched, however the large quantities of discovered coins enable for a more reliable chronological frame (Furlani 1969; Maselli Scotti 1992). The naturally protected fortification at an elevation of 150 m a.s.l. stood alongside the road that connected the two important roads towards Italy: the northern that lead along the Vipava Valley and the southern coastal road (Maselli Scotti 1992, 372). The prehistoric rampart was reinforced in Late Antiquity (Figs. 1.24,1.25). The trenches have shown that the wall was between 3 and 3.5 metres thick and that it could be dated into the beginning of the 5th century. The fortification was inhabited from the 3rd century onwards but the finds are more abundant from the transition between the 4th and 5th century on. 800 Late Roman coins from that period - maybe remains of a hoard - were discovered. The assumption that the find belongs to the time of Attila's invasion in 452 cannot be excluded (Maselli Scotti 1992, 372). Similar to the arrowheads the ceramic finds could also be dated after the mid 5th century. Maselli Scotti even assumed that this could be the remains of a fortification with a Byzantine garrison (Maselli Scotti 1992, 373). Late Roman settlements dated with the use of coins in the Posočje area as well as in the broader Eastern Alpine area go up to the third decade of the 5th century. The discontinuation of the coin circulation and the lack of other chronologically definable objects do currently not permit us to precisely define the time in which these settlements and fortifications were abandoned. Thus special attention is raised by the site in the vicinity of Robič, where a 5th century post is indicated. On Molida east of village Robič in the stone deposit of a landslide that broke from the northeast slope of Matajur, a solidus of Valentinian III was found by chance in 1892. The coin was minted between 425 and 455 (Župančič 1991, 167). During the railway construction a wall and some pottery fragments were supposedly found on this location Sl. 1.24: Gradina pri Doberdobu (Castellazzo di Doberdo). Pogled na utrdbo z južne strani. Fig. 1.24: Gradina near Doberdob (Castellazzo di Doberdo). A view of the fort from the south. Sl. 1.25: Gradina pri Doberdobu (Castellazzo di Doberdo). Načrt poznoantične utrdbe (Furlani 1969, sl. 65). Fig. 1.25: Gradina near Doberdob (Castellazzo di Doberdo). Site plan of the Late Antique fort (Furlani 1969, Fig. 65). 1992). Naravno zavarovana utrdba na n. m. v. 150 m je ležala ob cesti, ki je povezovala obe pomembni vpadnici proti Italiji: severno, ki je tekla po Vipavski dolini, in južno obalno cesto (Maselli Scotti 1992, 372). Prazgodovinski okop so ponovno utrdili v pozni antiki (sl. 1.24, 1.25). Sonde so pokazale, da je obzidje debelo 3-3,5 m in (Osmuk 1985e, 297). The poem Molida by J. Lavrenčič mentions a Roman sword and the story told by the locals speaks about a town that was covered by a landslide from Matajur (because of the sins of its inhabitants). This data indicates a potential Late Antiquity site that cannot be precisely ascertained. ga je mogoče datirati v začetek 5. st. Utrdba je bila obljudena od 3. st. dalje, zgoščene najdbe pa so s prehoda med 4. in 5. st. Našli so 800 poznorimskih novcev - morda depo - iz istega časa. Ni izključena domneva, da sodi najdba v čas Atilovega vpada leta 452 (Maselli Scotti 1992, 372). Keramične najdbe je podobno kot puščične osti mogoče datirati tudi v čas po sredini 5. st. Maselli scottijeva domneva v utrdbi celo bizantinsko posadko (Maselli Scotti 1992, 373). Vse z novci dobro datirane poznorimske naselbine v Posočju in tudi na širšem območju vzhodnih Alp segajo v tretje desetletje 5. st. Prenehanje novčnega obtoka in pomanjkanje drugih kronološko občutljivih predmetov nam za sedaj onemogočata natančneje opredeliti čas opustitve teh naselbin in utrdb. Zato zbuja posebno pozornost najdišče v bližini Robiča, kjer se nakazuje postojanka iz 5. st. Leta 1892 so namreč na Molidi, ledini vzhodno od naselja Robič, v kamnitem nanosu plazu, ki se je utrgal s severovzhodnega pobočja Matajurja, po naključju našli solidus Valentinijana III., ki je bil kovan med letoma 425-455 (Župančič 1991, 167). Pri gradnji železnice naj bi na tem mestu naleteli na zidovje in lončenino (Osmuk 1985e, 297). V pesmi J. Lavrenčiča Molida pa je omenjen rimski meč in tudi pripoved domačinov pozna mesto, ki ga je zaradi grešnosti njegovih prebivalcev zasul plaz z Matajurja. Ti podatki nakazujejo potencialno poznoantično najdišče, ki pa ga podrobneje ni mogoče označiti. Podobno kot opustitve poznorimskih postojank v prvi polovici 5. st., doslej ni bilo mogoče v vzhodno-alpskem svetu natančneje datirati niti nastanka zadnjih oblik poznoantičnih utrjenih višinskih naselbin, ki so najznačilnejši izraz transformacije poznoantičnega sveta (Ciglenečki 1987a, 114-116). Raziskave na Tonovcovem gradu so sicer dokazale obstoj velike utrjene naselbine s številnimi stanovanjskimi zgradbami in cerkvami po sredini 5. st., pri čemer pa je verjetnejša zadnja tretjina 5. st. (glej pogl. 2.2). To datacijo posredno potrjujejo primerjave z drugimi sorodnimi naselbinami v vzhodno-alpskem prostoru, ki so bolje datirane z že raziskanimi grobišči (Rifnik, Kranj, Teurnia idr.). V zadnjo skupino poznoantičnih utrjenih naselbin v Posočju bi poleg Tonovcovega gradu smeli uvrstiti še dve manjši postojanki, ki pomembno dopolnjujeta poselitveno sliko. Površinsko zelo izrazit je Gradec pri Drežnici, ki leži v neposredni bližini Tonovcovega gradu na levem bregu soče na severnem odrastku hriba ozben na robu Drežniškega kota (Osmuk 1985b; Ciglenečki 1997b, 25). Dviguje se strmo nad kanjonom potoka Kozjeka, na drugih straneh pa sega le nekaj metrov nad okolico (sl. 1.26). Približno 80 x 20 m velika površina hriba je bila na mestih, ki niso bila naravno zavarovana, to je predvsem na južni in vzhodni strani, dobro utrjena z okopom. Ker arheološke raziskave na hribu niso bile opravljene, ni mogoče reči, ali se v dobro vidnem okopu skriva kombi- It has been so far impossible to precisely date the abandonment of the Late Roman sites during the first half of the 5th century as well as the foundation of the Late Antique hilltop settlements that represent the most characteristic expression of the transformation of the Late Antique world in the Eastern Alps (Ciglenečki 1987a, 114-116). The research at Tonovcov grad has proven the existence of a large fortified settlement with numerous living quarters and churches after the middle of the 5th century, but it is more likely that the settlement was founded in the last third of the 5th century (see chapter 2.2). This data is indirectly confirmed by the comparisons to other similar settlements in the Eastern Alpine area that have been more precisely dated with the use of the explored cemeteries (Rifnik, Kranj, Teurnia, etc.). Alongside Tonovcov grad two smaller posts that importantly add to the settlement pattern should be added to the last group of Late Antique fortified settlements in the Posočje region. Gradec near Drežnica lies in the vicinity of Tonovcov grad - on the left side of the river Soča and on the northern slope of the hill Ozben which stands on the edge of Drežniški kot (Osmuk 1985b; Ciglenečki 1997b, 25). On one side it rises sharply above the stream of Kozjek, while on the remaining sides it raises above its surroundings by merely a few metres (Fig. 1.26). In the spots where it was not naturally protected, i.e. especially on the southern and eastern side, the area of the hill measuring approximately 80 x 20 m was fortified with a rampart. Because no archaeological research was performed on the hill, it is impossible to say whether a combination of earthwork and quarry stone or even a wall construction with mortar is hiding beneath it. As much as it was possible to ascertain merely by observing the surface, the first option seems to be the likeliest. The surface in the interior is artificially levelled out and shows a number of variously wide terraces that gradually descend towards the northwest (Fig. 1.27). Some modest surface finds (especially pottery) place the settlement into the Late Antique period. The lack of surface traces that would indicate the presence of permanent building forms in the extremely small area of the post as well as the modest surface finds indicate a naturally well protected short term refuge. We should consider the possibility of a fort that, in the case of need, protected or closed the route from the Soča Valley towards Drežnica, for this was one of the most important entries into Drežniški kot. Taking into account the vastness of the area under the mountain of Krn we could assume scattered settlements or individual homesteads of the autochthonous population who retreated to this position during the Late Antiquity. Undoubtedly the inhabitants who used this fortified settlement saw their centre in the nearby settlement of Tonovcov grad, which is just over 1 kilometre away. This fortified settlement indicates an autochthonous population's enclave in Drežniški kot, Sl. 1.26: Gradec pri Drežnici. Pogled na najdišče z zahodne strani. Fig. 1.26: Gradec near Drežnica. A view of the site from the west. Sl. 1.27: Gradec pri Drežnici. Poskus zarisa obsega naselbine po lidarskem posnetku (© ZRC SAZU). Fig. 1.27: Gradec near Drežnica. Mapping the settlement on a lidar data visualization (© ZRC SAZU). nacija ilovnatega nasipa in kamnitih lomljencev ali celo z malto zidana konstrukcija. Kolikor je mogoče soditi zgolj po površinskem ogledu, se zdi najverjetnejša prva možnost. Površina v notranjosti je umetno zravnana in kaže več različno širokih teras, ki postopoma padajo proti severozahodu (sl. 1.27). Nekatere skromne površinske najdbe (predvsem keramike) naselje opredeljujejo v poznoantično obdobje. Odsotnost sledov na površini, ki bi nakazovali prisotnost trajnih oblik zgradb na izredno majhni površini postojanke, kot tudi skromne površinske najdbe nakazujejo predvsem naravno dobro which has been additionally confirmed by the finds discovered in the studies of the folklore, names and other research (Kuret 1972; 1989; Volarič 2008). Gradec near Logje - the third settlement that can be reliably dated into Late Antiquity - is located in Breginjski kot next to the Italian border (Osmuk 1985d; Ciglenečki 1997b, 25). The hill (446 m a.s.l.), which is naturally well protected by slopes and on the eastern side by a vertical rock face, is located at the confluence of the river Nadiža and the stream Legrado (Fig. 1.28). It can be accessed only from the southern side, and even then a long saddle and a steep slope need to be passed. The width at the top of the exposed ridge rarely exceeds 20 metres and the ridge measures approximately 120 metres in length. Minute traces where the terrain has been levelled out are visible, and for two points it can be assumed that they created a space for improvised wooden constructions. Due to its naturally protected position it remains an excellent refuge in a region which is sparsely inhabited even today. There are no traces of a rampart at the top, however it was most likely also not needed (Fig. 1.29). A recently discovered fibula (Osmuk 1999, 64-66; 2001, 47) proves that it was inhabited in the 6th century, thus it could be assumed to be a Late Antique refuge, similar to the one at Gradec near Drežnica. This purpose is indirectly indicated by extremely modest surface finds. It seems that the inhabitants were - due to the remoteness of the area from the transport route connecting Cividale del Friuli and Carinthia - safe and that the refuge in Gradec was used only in extreme cases. There are no settlements in the central part of the Posočje region that could be reliably dated into the final phase of the late Late Antiquity. There is a high zavarovano kratkotrajno pribežališče. Pomisliti bi smeli tudi na utrdbo, ki je v primeru potrebe varovala oziroma zapirala pot iz doline Soče proti Drežnici, saj je tu eden pomembnejših dohodov v Drežniški kot. Glede na veliko prostranost območja pod Krnom bi smeli domnevati tu raztresene zaselke ali posamezne domačije staroselskega prebivalstva, ki se je sem umaknilo v času pozne antike. Nedvomno pa so prebivalci, ki so uporabljali to utrjeno naselje, svoje središče videli na bližnjem Tonovcovem gradu, ki je v zračni črti oddaljen le dober kilometer. Ta utrjena naselbina posredno nakazuje staroselsko enklavo v Drežniškem kotu, kar dodatno potrjujejo tudi izsledki folklorističnih, imenskih in drugih raziskav (Kuret 1972; 1989; Volarič 2008). Tretja zanesljiva poznoantična naselbina Gradec pri Logjeh leži v Breginjskem kotu tik ob italijanski meji (Osmuk 1985d; Ciglenečki 1997b, 25). Hrib (višine 446 m n. m.), ki je že po naravi odlično zavarovan s strminami in na vzhodni strani celo z navpično skalno steno, leži v sotočju Nadiže in potoka Legrado (sl. 1.28). Dostop nanj je mogoč le z južne smeri čez daljše sedlo in strmo pobočje. Širina na vrhu zelo izpostavljenega grebena je le redko kje več kot 20 m, v dolžino pa meri okoli 120 m. Vidni so neznatni sledovi izravnav terena, na dveh mestih je mogoče slutiti prostor za improvizirane lesene konstrukcije. Zaradi svoje naravno zavarovane lege je v še danes redko poseljenem svetu odlično zatočišče. okopa na vrhu ni videti, v večjem delu tudi ni bil potreben (sl. 1.29). Pred nedavnim najdena fibula (Osmuk 1999, 64-66; 2001, 47) dokazuje poseljenost v 6. st., zato bi, podobno kot na Gradcu pri Drežnici, tudi tu smeli domnevati pribežališče iz poznoantičnega obdobja. To namembnost posredno nakazujejo izredno skromne površinske najdbe. Zdi se, da so bili prebivalci zaradi odmaknjenosti celotnega območja od prometne žile Cedad-Koroška dovolj varni in so zato pribežališče na Gradcu uporabljali le v izjemnih primerih. V srednjem delu Posočja dokazanih naselbin iz sklepne faze poznoantičnega obdobja ne poznamo. Zelo verjetna se zdi naselbina na že prej opisani Sv. Katarini nad Novo Gorico (sl. 1.21,1.22), kjer pa sondiranja niso nedvoumno potrdila obstoja te faze (Svoljšak 1990, 43; Osmuk 1997b). Vendar bi ga smeli upravičeno pričakovati: takrat že obstoječe močno obzidje, ki je bilo utrjeno z vsaj štirimi izstopajočimi stolpi, je pravzaprav značilnost prav posebnega tipa poznih postojank, kakršne smo dobro spoznali pri raziskovanju Korinjskega hriba nad Velikim Korinjem (Ciglenečki 1985). Tudi Rutarjeva omemba s "cementom zidanih grobov" v povezavi s postojanko na Sv. Katarini bi utegnila kazati na naslednji značilen element poznih strateško umeščenih utrdb: v doslej raziskanih so bile kot največja značilnost poleg močno utrjenega obzidja s stolpi tudi manjše zgodnjekr-ščanske cerkve, ob katerih so zelo pogoste zidane grobnice (Korinjski hrib, Zidani gaber). Zanimiv je tudi pri Bavčarju objavljen podatek o leta 1600 najdenih zlatih Sl. 1.28: Gradec pri Logjeh. Pogled na najdišče z južne strani. Fig. 1.28: Gradec near Logje. A view of the site from the south. Sl. 1.29: Gradec pri Logjeh. Poskus zarisa obsega naselbine po lidarskem posnetku (© ZRC SAZU). Fig. 1.29: Gradec near Logje. Mapping the settlement on a lidar data visualization (© ZRC SAZU). kovancih, ki se nanaša na Sv. Katarino (Svoljšak 1990, 35-37). ti bi, podobno kot smo opozorili pri Molidi, morda kazali na poznoantično zakladno najdbo, kakršne so bile odkrite v mnogih utrdbah na strateških mestih. S sondiranji pridobljene skope najdbe teh domnev še ne potrjujejo, čeprav je med grobo keramiko nekaj fragmentov, ki imajo dobre analogije prav v gradivu najpoznejše faze poznoantičnih višinskih postojank v Sloveniji (tonovcov grad, Ajdna). težava pri tovrstni določitvi je predvsem v tem, da je bila utrdba poseljena še pozneje v slovanskem obdobju, ko so uporabljali zelo sorodno keramično gradivo. Nekoliko zunaj obravnavanega območja, že globoko v Vipavski dolini, pa leži še ena izredno dobro ohranjena poznoantična višinska postojanka - Sv. Pavel nad Vrto-vinom, ki odlično dopolnjuje podobo zadnjih antičnih naselbin (Svoljšak 1985). Njene značilnosti so daleč po dolini vidna, izpostavljena lega, še danes visoko ohranjeni poznoantični zidovi in delitev naselbine v dva dela (sl. 1.30,1.31). Mogočna skalna utrdba leži na južnem pobočju Čavna, na n. m. v. 525 m in je nad okolico dvignjena 10-30 m. Zgrajena je bila na prepadnem skalnem osamel-cu, na vrhu katerega je pribl. 500 m dolg in 70-130 m širok plato. Na treh straneh je naravno zavarovana s skalnimi stenami. Nekoliko bolj dostopna je severna stran, kjer je bil med skalami prirejen tudi vhod v naselbino. Manjša sondiranja žal niso razkrila zidane arhitekture v takšni meri, da bi jo lahko primerjali z drugimi sočasnimi objekti. Vendar že njena nenavadna ureditev teras kaže, da je bila naselbina zasnovana drugače. Postojanka je z 2,5 m širokim prečnim zidom razdeljena na dva dela. Nekoliko slabše zavarovani severni del je bil intenzivneje poseljen, južni pa kaže le občasno poselitev, zato je najbrž rabil za pribežališče. ob robu skal je še dobro ohranjen obrambni zid, ki je bil speljan tudi nad prepadnimi stenami. Širok je 1 m, ob njem je vodila cesta. Pri sondiranjih leta 1966 sta D. Svoljšak in P. Petru ugotovila stavbne ostaline na treh mestih. južno od cerkve sta našla manjšo stavbo trapezoidnega tlorisa, ki je bila zgrajena iz slabega zidu, debelega 55 cm. Del večje in boljše grajene stavbe je bil najden tik za današnjo cerkvijo. Ostanek zidu tretje stavbe so izkopali na terasah, ki ležijo na skrajnem severu utrdbe in so bile - sodeč po vidni izoblikovanosti terena na površini - osrednji prostor poselitve. Posebno zanimiv je vodni stolp, ki leži tik pod naselbino in je velik 6,7 x 10 m (Petru 1972, 359-361). S severa se v njegov bazen spuščajo stopnice, ki so deloma vsekane v živo skalo. Med najdbami je najštevilneje zastopana groba kuhinjska keramika, ki najdišče časovno dobro umešča v 5. in 6. st. Arheološko doslej nepreverjeni so ostanki domnevne naselbine Na kapli v Podmelcu (sl. 1.32). V neposredni bližini manjšega grobišča, ki ga pridatki datirajo v 6. st., so namreč vidni obrambni okopi, ki jih je v skico najdišča vrisal V. Šribar (Šribar 1967). teh Sl. 1.30: sv. pavel nad Vrtovinom. pogled na skalni plato z južne strani. Fig. 1.30: sv. pavel above Vrtovin. A view of the rocky plateau from the south. probability that a settlement existed on the previously mentioned Sv. Katarina above Nova Gorica (Figs. 1.21, 1.22), even though the trenching did not confirm its existence in this phase (svoljšak 1990, 43; Qsmuk 1997b). However, this phase is to be expected: there was a strong wall fortified by at least four protruding towers at the time, and this is a characteristic of a special type of late posts that have been well researched and documented during the research of Korinjski hrib above Veliki Korinj (Ciglenečki 1985). Even Rutar's mention of 'graves built with the use of concrete' in relation to the post on sv. Katarina could indicate the next typical element of the late strategically placed forts: in the ones researched so far the most common characteristic (apart from the strongly fortified walls and towers) is represented by small Early Christian churches that were often accompanied by tombs (Korinjski hrib, zidani gaber). Interesting is also the data that was published in Bavčar's book and deals with the gold coins linked to sv. Katarina that were discovered in 1600 (svoljšak 1990, 35-37). similar to the ones found at Molida these coins could indicate a Late Antique hoard as is the case at numerous other fortifications on strategic positions. The modest finds discovered through trenching do not confirm these assumptions, even though a few pottery fragments that can be compared to the coarse wares from the latest phase Sl. 1.31: Sv. Pavel nad Vrtovinom. Načrt utrdbe (po Svoljšak, Knific 1976, sl. 7). Fig. 1.31: Sv. Pavel above Vrtovin. Site plan of the fortification (after Svoljšak, Knific 1976, Fig. 7). okopov ni mogoče prepričljivo umestiti v poznoantično obdobje brez arheoloških raziskav, saj se v marsičem razlikujejo od drugih doslej zanesljivo znanih najdišč. Nerazjasnjen je tudi pomen dveh grobelj, v katerih je Šribar opazil suhi zid, domačini pa so na ta dva objekta navezovali pripoved o obstoju kapele, ki naj bi bila razlog za poimenovanje najdišča. Bolj izpovedno je grobišče Na Bukovčevem brdu nedaleč vstran, kjer je bilo pri sondiranjih leta 1963 in 1965 izkopanih 5 skeletnih grobov (Šribar 1967, 379-386). Pridane pasne spone so takrat dovoljevale datacijo grobišča v 6. st. Na večji obseg grobišča kažejo omembe grobov, ki so jih izkopali of Slovenian Late Antique hilltop posts (Tonovcov grad, Ajdna) were discovered. The problem with this definition is that the fort was inhabited also later in the Slav period when very similar pottery was used. Somewhat outside the discussed area, deep in the Vipava Valley, lies another extremely well preserved Late Antique hilltop post - Sv. Pavel above Vrtovin -which provides an excellent addition to the image of the last settlements of the Antiquity (Svoljšak 1985). Its characteristics are an exposed and far seen position, even today high up preserved Late Antique walls and the division of the settlement into two parts (Figs. 1.30; 1.31). The mighty rock fort stands on the south slope of Čaven, at 525 m a.s.l. and is about 10-30 m above its surroundings. It was built on an isolated rock hill with an approximately 500 m long and a 70-130 m wide plateau. It is naturally protected by rock faces on three sides. Somewhat more accessible is the northern side, where the entrance into the settlement was created amongst the rocks. Unfortunately the modest trench interventions did not reveal any stone architecture to the extent where it could be compared to other buildings from the time. However, already the unusual organisation of its terraces shows that the settlement was conceived differently. The post was divided into two parts by a 2.5 metre wide wall. The somewhat poorer protected northern part was more intensively inhabited, while the southern part shows only occasional settlement, thus it was most likely used as a refuge. At the edge of the rocks stands a well preserved defensive wall that ran also above the precipitous rock face. It is 1 metre wide and a road used to run alongside it. During the 1966 trenching D. Svoljšak and P. Petru ascertained building remains in three locations. South of the today's church they found a smaller building with a trapezoid ground plan that was built with a weak wall, only 55 cm thick. A part of the larger and better constructed building was found just behind today's church. The wall remains of the third building were excavated on the terraces that can be found on the northernmost part of the fort and represented - according to the layout of the terrain visible on the surface - the central space of the settlement. Especially interesting is the water tower that stood below the settlement and measured 6.7 x 10 m (Petru 1972, 359-361). Stairs, some of which are carved into the bedrock, lead into its pool from the north. Coarse kitchenware is the most common amongst the finds and it dates the site into the 5th and 6th centuries. Archeologically still unverified are the remains of the assumed settlement Na Kapli in Podmelec (Fig. 1.32). In the vicinity of a smaller graveyard, dated into the 6th century according to the grave goods, ramparts can be seen. These ramparts were drawn into the sketch of the site by V. Šribar (Šribar 1967). Without further archaeological research these ramparts cannot be reliably dated to the Late Antique period, for they differ from Sl. 1.32: Podmelec. Skica domnevne naselbine (Šribar 1967, sl. 1). Fig. 1.32: Podmelec. A sketch of the presumed settlement (Šribar 1967, Fig. 1). že pred prvo svetovno vojno (Šribar 1967, 379; Svoljšak 1975, 210; Vuga 1979). Pasni sponi iz grobov sta bili pred kratkim konservirani in kažeta značilnosti poznega oblikovanja iz časa okoli leta 600 (Knific 2004, 13). Ostale poznoantične naselbine lahko, podobno kot v Podmelcu, posredno nakažemo le s po naključju znanimi manjšimi grobišči. Takšna je nekropola Na Mirih v Ljubinju neposredno ob stari poti, ki je vodila iz Tolmina čez Podmelec v dolino Bače. Manjša sondiranja Timoteja Knifica na mestu, kjer so že prej našli grobove (omenjajo zlat prstan), so pokazala na obstoj manjšega grobišča, ki ga je mogoče umestiti v drugo polovico 6. st. (Cvitkovič 1999, 40; Knific 2010). Naslednje grobišče na Ledinah v Novi Gorici je znano z območja, ki je bilo gosteje obljudeno v pozni other sites that have been reliably dated. The meaning of the two cairns in which Šribar noticed a dry wall is also unclear. The local inhabitants linked to them the story of the chapel, and supposedly this was the reason behind the name of the site. More telling is the nearby graveyard on Bukovčevo brdo, where five skeletal graves were discovered during the 1963 and 1965 trenching (Šribar 1967, 379-386). The belt buckles found in the graves dated the cemetery into the 6th century. The mentions of graves that were excavated prior to World War I indicate that the cemetery was probably larger (Šribar 1967, 379; Svoljšak 1975, 210; Vuga 1979). The two belt buckles from the graves have been recently conserved and show form characteristics that can be dated to around the year 600 (Knific 2004, 13). antiki (glej karto najdišč pri Knific, Svoljšak 1984, 279, sl. 1). V ostanke poznorimskih zgradb je bilo vkopanih pet skeletnih grobov z značilnimi, s kamni obdanimi grobnimi jamami. Pridatki skeletnih grobov so bili skromni, vendar kažejo izrazit staroselski značaj (Osmuk 1987b; Svoljšak 1987). Pripadajoče naselje ni bilo ugotovljeno, glede na lego grobišča na bolje zavarovanem območju pa bi smeli pomisliti na manjši ravninski zaselek v neposredni bližini, morda prav tako umeščen v ostanke starejših rimskih zgradb. Izredno pomembno je, žal v veliki meri uničeno, langobardsko grobišče iz časa langobardske selitve v Italijo v Biljah. Oprema, ki nakazuje pokop več langobard-skih vojščakov, opozarja na bližnjo neodkrito postojanko tik ob pomembni vpadnici v Italijo, ki bi jo morda smeli iskati celo v bližnjih podrobneje neraziskanih rimskih ruševinah (Osmuk 1978). Po temeljiti konservaciji takrat najdenih predmetov so na orožju ugotovili še nekatere dodatne podrobnosti (Knific 2004, 13). Nenavadna je najdba žganega germanskega groba z najdišča Rodne pri Tolminu. Iz pisma Karla Moserja z dne 26. 7. 1895 Naravoslovnemu muzeju na Dunaju je mogoče razbrati, da je bil na omenjeni ledini leta 1890 odkopan žgan grob, ki je ležal kar 1,80 m globoko (sl. 1.33). 9 V njem so našli bronast umbo in železen meč z - kot je mogoče razbrati z Moserjeve skice - bronasto nožnico. zraven je bil najden tudi zlat prstan, ki pa se je izgubil. Moser je ob ogledu najdišča sklepal na posamični grob. Usoda najdb je neznana, Moser je v pismu nakazal pripravljenost, da najdbe pošlje v dunajski muzej. Moserjevo pisno poročilo zastavlja nekaj problemov: grob naj bi ležal na desni strani Soče, pol ure hoda proti Mostu na Soči. Jan Cvitkovič je najdbo lociral na rob 9 Na grob nas je že pred leti prijazno opozoril Dragan Božič, za kar se mu na tem mestu zahvaljujemo. Sl. 1.33: Rodne pri Tolminu. Moserjeva risba predmetov iz germanskega groba (pismo K. Moserja z dne 26. 7. 1895. Hrani Naravoslovni muzej Dunaj). Fig. 1.33: Rodne near Tolmin. Moser's drawing of the objects from the Germanic grave (Karl Moser's letter, dated 26th July 1895. Kept by the Naturhistorisches Museum in Vienna). Similar as is the case in Podmelec, other Late Antique settlements can only be indirectly indicated with the accidentally discovered small cemeteries. Such is for instance the cemetery Na Mirih in Ljubinj which is located on the old route that led from Tolmin across Podmelec and into the valley of Bača. Small trenching interventions carried out by Timotej Knific on the location at which graves have been previously found (a gold ring is mentioned), have shown the existence of a small cemetery that can be dated into the second half of the 6th century (Cvitkovič 1999, 40; Knific 2010). The next cemetery on Ledine in Nova Gorica is known from the area that was more densely populated during the Late Antiquity (see map of sites in Knific, Svoljšak 1984, 279, Fig. 1). Five skeletal graves with characteristic grave pits encircled with stones were dug into the remains of Late Roman buildings. The grave goods were modest; however they are characteristic of the autochthonous population (Osmuk 1987b; Svoljšak 1987). No accompanying settlement was ascertained, however taking into account that the cemetery was located in a secure Italic area we could consider the existence of a smaller unprotected plain settlement in the vicinity, possibly positioned on the remains of older Roman buildings. Extremely important, however unfortunately mainly destroyed, is the Lombard cemetery in Bilje that can be dated to the time the Lombards were moving to Italy. The grave goods indicate the burials of a number of Lombard soldiers and lead us to believe into the existence of a yet undiscovered post in the vicinity. This post stood alongside the important route to Italy and could have been located on the nearby not so well researched Roman ruins (Osmuk 1978). Following thorough conservation efforts of the objects discovered at the time additional details were discovered on the weapons (Knific 2004, 13). Unusual is the discovery of the Germanic cremation burial at the site Rodne near Tolmin. In Karl Moser's letter to the Naturhistorisches Museum in Vienna, dated 26th July 1895, he mentions that a 1.80 metre deep cremation burial was excavated there in 18909 (Fig. 1.33). The grave revealed a bronze shield boss and an iron sword with - as depicted in Moser's sketch - a bronze sheath. A gold ring was discovered in the grave, however it was lost through time. When viewing the site Moser assumed it was an isolated grave. The fate of the finds is unknown, however in the letter Moser indicated his willingness to send the finds to the museum in Vienna. Moser's report poses a few problems: the grave was supposedly positioned on the right bank of the Soča river, half an hour by foot towards Most na Soči. Jan Cvitkovič located the site on the edge of Tolmin, where the settlement mentioned in the letter is located (Cvitkovič 1999, 42). Contrary to Moser's data Rodne is situated on the left bank of Soča as well as on 9 Dragan Božič told us about this grave years ago, and we would like to take this opportunity to thank him. Tolmina, kjer leži v pismu omenjeni zaselek. V nasprotju z Moserjevimi podatki pa ta ledina dejansko leži na levi strani Soče in celo na levi strani Tolminščice, ki se nekoliko južneje od tod izliva v Sočo (Cvitkovič 1999, 42). Najdišče tudi ni oddaljeno pol ure hoje od Tolmina, ampak le slab kilometer od njegovega središča. Zato pa se sklada podatek, da leži najdišče ob cesti za Sveto Lucijo (Most na Soči) in da je blizu obrežja reke, kjer so lahko pridobivali v pismu omenjeni gramoz. Cvitkovič datira grob glede na ustno mnenje A. Pleterskega v 7. st. Podrobnejši ogled Moserjeve skice pokaže, da gre nedvomno za moški grob s pridano spatho in umbom ter zlatim prstanom. Zaradi približno skiciranih predmetov ni možna njuna natančnejša opredelitev, analogije imata v širšem geografskem prostoru Barbarika kot tudi v mejah imperija na današnjem madžarskem in avstrijskem prostoru in ju je mogoče datirati v 6. st. Pomenljiva je tudi globina groba: podobna je bila ugotovljena pri najmanjši skupini najpomembnejših germanskih grobov v Dravljah (Slabe 1975, 34-36) in na Lajhu v Kranju (Stare 1980, 89; Odar 2006, 246). Nenavadno je predvsem dejstvo, da je grob žgan, kar ga razlikuje od doslej odkritih germanskih grobov na slovenskem ozemlju. Najbližje žgane grobove je mogoče zaslediti v Kajdacsu na Madžarskem, kjer so bili po mnenju I. Bone poleg langobardskih vojščakov pokopani tudi zavezniški Sasi, ki so ohranili star običaj pokopavanja (Bona 1970-1971, 49). Možnost, da je bil v Rodnah pokopan langobardski zaveznik, bi utegnilo dati pomemben namig glede smeri poti, po kateri so se Langobardi s svojimi zavezniki selili v Italijo. Če pritegnemo Moserjevo opažanje, da gre za posamičen grob in dejstvo, da so stare poti potekale po levi strani Soče med Tolminom in Mostom na Soči, bi smeli upravičeno pomisliti na bojevnika, ki je omagal na dolgi poti iz Panonije v Italijo, tik pred obljubljeno deželo. To je seveda le ena izmed možnosti, vendar ne povsem neverjetna glede na dejstvo, da žgani germanski grobovi v Italiji niso bili odkriti. Zato bi smeli upravičeno domnevati pokop v času pohoda Langobardov leta 568 ali pa kmalu zatem, ko so se Langobardi v tem prostoru že ustalili in za zavarovanje območja postavili manjše posadke, kot se je to potrdilo že na Tonovcovem gradu (primerjaj najdbo spathe in držaja ščita: Tonovcov grad. Najdbe, t. 11: 10 in 46: 9). Gosto razpredena mreža poznorimskih cest in tovornih poti, ki jo podrobneje predstavljamo v poglavju o poselitvi v poznorimskem obdobju, je bila uporabljena pri prehodu Langobardov iz Panonije v Italijo (glej pogl. 5.1). Med redkimi sledovi teh premikov bi utegnil biti prav ta grob eden izmed odločilnih kazalnikov (Werner 1962, 121-130). Skromno, a pestro poselitveno podobo pozno-antičnega obdobja dopolnjujejo posamezne drobne poznoantične najdbe, ki jih je težko povezati s konkretnimi naselbinskimi sledovi. Takšni sta predvsem zgodnjekrščanska oljenka na ledini Pod Čahlnami in poznoantična fibula z ledine Varda v Novi Gorici, ki the left bank of Tolminščica, which runs into Soča slightly further south from here. The site is also not half an hour by foot from Tolmin, but less than 1 kilometre from its centre. On the other hand the mention in the letter that the site is positioned alongside the road towards Sveta Lucija (Most na Soči) and that it is close to the river bank where gravel was dug holds true. Taking into account the opinion passed on by A. Pleterski Cvitkovič dates the grave into the 7th century. A detailed view of Moser's sketch shows that it was undoubtedly a male grave with a spatha, shield boss and a gold ring. Due to the rough sketches of the first two objects it is impossible for them to be precisely defined, however they have analogies in the broader geographical area of Barbaricum as well as within the Empire (on present day Hungarian and Austrian grounds) that can be dated into the 6th century. The depth of the grave is also important: a similar depth was ascertained in the smallest group of the most important Germanic graves in Dravlje (Slabe 1975, 34-36) and on Lajh in Kranj (Stare 1980, 89; Odar 2006, 246). Unusual is the fact that this is a cremation grave, which makes it different from the other Germanic graves that have been discovered in Slovenia. The closest cremation graves were discovered in Kajdacs, Hungary, where - according to I. Bona - not only the Lombard soldiers but also allied Saxons were buried, and they preserved the old burial ritual (Bona 1970-1971, 49). The possibility that a Lombard ally was buried in Rodne could provide an important hint as regards the direction of the route along which the Lombards and their allies moved towards Italy. If we agree with Moser that this was an isolated grave and take into account the fact that the old routes between Tolmin and Most na Soči ran on the left side of the Soča river, we could rightly assume that this was a warrior who failed to make the long journey from Pannonia to Italy, and died just prior to reaching the promised land. Of course, this is merely one of the possibilities; however, it becomes more probable if we take into account the fact that Germanic cremation graves have so far not yet been discovered in Italy. We could assume that the burial took place during the Lombard march in 568 or soon afterwards, when the Lombards settled in this area and positioned small garrisons that protected the area, as has been confirmed for Tonovcov grad (compare the spatha and the shield handle finds: Tonovcov grad. Finds, Pls. 11: 10 and 46: 9). When the Lombards were making their way from Pannonia to Italy they used the dense network of Late Roman roads and transport routes (see chapter 4.1). Amongst the rare traces of these movements this grave could represent a decisive indicator (Werner 1962, 121-130). In addition to the modest and diverse Late Antique settlement image we have the individual Late Antique finds that are hard to link with actual settlement traces. Such are for instance the early Christian oil lamp from Pod Čahlnami and the Late Antique fibula from Varda in Nova Gorica, both of which enrich the knowledge of the Sl. 1.34: Sv. Helena v PodbeJi. Pogled na naselbino z južne strani Fig. 1.34: Sv. Helena in Podbela. A view of the settlement from the south. Sl. 1.35: Sv. Helena v Podbeli. Lidarski posnetek površja (© ZRC SAZU) z opaznimi sledovi stavb (Kokalj, Zakšek, Oštir 2011, 265, sl. 1). Fig. 1.35: Sv. Helena in Podbela. Remains of structures visible on a lidar data visualization (© ZRC SAZU; Kokalj, Zakšek, Oštir 2011, 265, Fig. 1). bogatita vedenje o poselitvenem prostoru pod utrjeno postojanko na Sv. Katarini (Zavrtanik 1982). Lepo okrašena dvoramna fibula je bila najdena tudi v območju Lokavca na Banjšicah, na območju torej, kjer je že F. Truhlar glede na odmaknjeno lego in predslovanske toponime domneval poznoantično enklavo (Knific 2004, 12; Truhlar 1976). Prav tako je nekaj najdišč, ki jih z dosedanjimi raziskavami in obhodi ni bilo mogoče povsem zanesljivo settlement area under the fortified post on Sv. Katarina (Zavrtanik 1982). A nicely decorated equal-arm fibula was found also in the area of Lokavec on Banjsice, i.e. in an area for which F. Truhlar assumed a Late Antique enclave - due to the remote position and the pre-Slav toponyms (Knific 2004, 12; Truhlar 1976). The research carried out so far failed to provide precise data that would place certain sites into this period: however, they should be mentioned because of certain umestiti v to obdobje: zaradi nekaterih najdb in njihove lege ter oblikovanosti površja pa se vendarle zdi smiselno opozoriti nanje. Takšna je Sv. Helena v Podbeli, od koder je bila pred kratkim pridobljena tudi železna triroba puščična ost (Knific 2004, 9). Zravnan plato na naravno odlično zavarovanem hribu nakazuje na površju sledove stavb. Ti so se posebej dobro pokazali po uporabi metode zračnega laserskega skeniranja (lidar), ki smo ga izvedli leta 2007 (sl. 1.34, 1.35): pri tem je bilo mogoče na površini zaznati več obrisov stavb, ki bi utegnile soditi v poznoantično obdobje (Kokalj et al. 2011). Povsem vprašljiv je obstoj poznoantične postojanke na Kozlovem robu nad Tolminom, kjer naj bi pri izkopavanjih leta 1964 našli poznoantično žlico (Vuga 1970, 172) oziroma "nekaj elementov, ki kažejo na poznoantično poselitev gradu" (Svoljšak 1970-1971, 154). Novejše raziskave srednjeveškega gradu tega doslej niso potrdile. Fragmenti poznoantične oziroma zgodnjesrednjeveške keramike naj bi bili najdeni tudi v Selih nad Podmelcem (Osmuk 1985f, 297), a je zaradi majhnih in slabo določljivih delcev keramike težko potrditi datacijo. Drugačna je situacija v Kanalu ob Soči, kjer so na obrežju soče našli značilno poznoantično trirobo puščično ost, že prej pa je bil v Gorenji vasi, ki leži v neposredni bližini, najden Konstantinov zlatnik (Žbona Trkman 1981; Kos 1988, 66, št. 29). Terenski pregledi doslej niso pojasnili lege naselbine. Glede na zgodnejšo varianto najdene puščične konice bi obe najdbi utegnili kazati na obstoj poznorimske postojanke ali tudi na naselje, ki je bilo zaradi pomembne strateške lege ob prehodu čez reko obljudeno ali celo utrjeno za daljši čas v obdobju pozne antike. Povsem neznatni in neizpovedni so arheološki sledovi s konca 6. stoletja in naslednjih dveh stoletij, ko naj bi - glede na sporočila pisnih virov - postopoma naseljevali Posočje slovanski prišleki (Štih 1999). Prav to obdobje je doslej najskrivnostnejše in si o njem ni mogoče ustvariti zanesljive podobe. Domnevati je mogoče, da so ponekod staroselci vztrajali še v 7. st., o čemer pričajo posamezni kosi nakita in orožja na Sv. Pavlu nad Vrtovinom, sv. Katarini nad Novo Gorico, v kobariškem območju pa na Tonovcovem gradu. Kdaj natanko je življenje v njih zamrlo, ni mogoče natančneje ugotoviti. Ker je bilo to območje sestavni del langobardske države, omenjene postojanke niso bile tako izpostavljene kot tiste v vzhodni in osrednji Sloveniji, nedvomno pa opuščene kmalu po koncu antičnega obdobja (Svoljšak, Knific 1976, 80-81). O tem zgovorno priča skromna prisotnost ostalin zgodnjesrednjeveških prebivalcev na Tonovcovem gradu, ki so izkoristili ruševine deloma že podrtih antičnih zgradb, da so si v njih uredili zasilna bivališča. finds, their position and terrain. Such is Sv. Helena in Podbela, where an iron trefoil arrowhead has been recently discovered (Knific 2004, 9). The levelled plateau on a naturally excellently protected hill indicates traces of buildings on its surface. These were especially visible when the laser scanning method - introduced in 2007 -was used (Figs. 1.34, 1.35): with this method a number of building outlines could be seen and they could belong to the Late Antique period (Kokalj et al. 2011). The Late Antique post on Kozlov rob above Tolmin is debatable, even though a Late Antique spoon was supposed to have been found at the site during the 1964 excavations (Vuga 1970, 172) as were 'certain elements that indicate a Late Antique settlement under the castle' (Svoljšak 1970-1971, 154). Until now this has not been confirmed by the newer research of the medieval castle. The fragments of the Late Antique or Early Medieval pottery were supposedly also found in Sela above Podmelec (Osmuk 1985f, 297), but due to the small and poorly definable fragments it is hard to confirm its date. In Kanal ob Soči the situation is different, for here a typical Late Antique trefoil arrowhead was found on the Soča riverbank, and even before that a gold coin of Constantine I was found in Gorenja vas, which lies in the immediate vicinity (Žbona Trkman 1981; Kos 1988, 66, No. 29). So far the terrain inspections have not explained the position of the settlement. Taking into account the earlier variant of the discovered arrowhead both finds could indicate the existence of a Late Roman post or even a settlement that was - due to its important strategic position at the river crossing - populated or even fortified for a longer time in Late Antiquity. Extremely minute and non-explanatory are the archaeological traces from the end of the 6th century and the following two centuries, when - as gathered from written sources - the Posočje area was gradually inhabited by the Slav newcomers (Štih 1999). So far this period remains the most mysterious and it is impossible to establish a clear image of it. It can be assumed that in certain places the autochthonous population persisted into the 7th century, which is indicated by the individual pieces of jewellery and weapons discovered at Sv. Pavel above Vrtovin, Sv. Katarina above Nova Gorica, and at Tonovcov grad in the Kobarid area. It is impossible to ascertain the precise time in which these settlements were abandoned. Because this area was a constituent part of the Lombard state, the aforementioned posts were not as exposed as those in eastern and central Slovenia and were undoubtedly abandoned soon after the end of Antiquity (Svoljšak, Knific 1976, 80-81). This is clearly shown by the modest presence of the remains from the later Early Medieval settlers who settled at Tonovcov grad and made best use of the ruins of the partially demolished Antique buildings in which they created their temporary dwellings. 1.6 mreža poti 1.6 path network Benjamin ŠTULAR 1.6.1 UVOD 1.6.1 INTRODUCTION v arheologiji se vedno znova srečujemo z vprašanji, povezanimi z vlogo posameznega najdišča v prostoru. Med najpogostejšimi je vprašanje poti, ki so določeno najdišče povezovale s pomembnimi kraji v prostoru. Pri tem se v arheologiji najpogosteje spopadamo s pomanjkanjem podatkov. V Sloveniji tako arheološko dokumentiranih poznorimskih ali poznoantičnih poti skorajda ne poznamo. Poznamo le primera poti v naselbini (Ajdovski gradec nad Vranjem: Petru, Ulbert 1975) oziroma na grobišču (Bled - Pristava: Pleterski 2008a, 75-95) in študijo o zgodnjesrednjeveških gospodarskih poteh v Blejskem kotu (Štular 2006a). Na drugi strani poznamo za to obdobje regionalne ali nadregionalne študije, pri katerih sklepamo o trasah poti na podlagi bližine naselbin in najdb (Ciglenečki 1985; 2007). Na tem mestu nas zanima vmesen, mikroregio-nalni pogled. Natančneje, gre za poti, ki so naselbino na Tonovcovem gradu povezovale z okolico, predvsem s sočasnimi centralnimi kraji (npr. Kranj, Čedad). Študijsko območje smo izbrali tako, da smo zajeli vse tri verjetne dostope do Tonovcovega gradu z juga in zahoda, torej dolino reke Soče južno od Kobarida ter dva kraka doline Nadiže od izvira do prehoda v Furlansko nižino. Na severu smo območje zamejili s križiščem, na katerem se razcepijo poti na različne alpske prelaze (sl. 1.36). Ker neposrednih arheoloških podatkov o poteh nimamo, se lahko opremo le na historično izpričane poti in na prostorske analize z daljinskim zaznavanjem ter geografskimi informacijskimi sistemi (GIS). 1.6.2 ZGODOVINSKI VIRI Najstarejši relevanten podatek, izpričan v zgodovinskih virih, se nanaša na t. i. Bovško pot, ki je v srednjem veku povezovala Furlanijo in Koroško. Približen potek poti je znan: čez Čedad je pot vodila proti severu po dolini Nadiže. Pri kraju Robič je zapustila Nadižo in se usmerila proti vzhodu, pri Kobaridu dosegla Sočo in ob njej potekala proti Bovcu ter dalje čez Predel in In archaeology we constantly deal with issues that are linked to the role of a site in its surroundings. Amongst the most common is the issue of paths that linked a certain site to the important places in its surroundings. However, while performing this archaeology (as a scientific field) usually does not have sufficient data at its disposal. For instance, there are almost no documented Late Roman or Late Antique paths in Slovenia. The only excavated ones are the paths within the settlement (Ajdovski Gradec above Vranje: Petru, Ulbert 1975) or the burial site (Bled - Pristava: Pleterski 2008a, 75-95). There is also the study on the Early Medieval economic paths in Blejski kot (Štular 2006a). On the other hand certain regional and interregional studies assume that the paths and roads ran close to the settlements (Ciglenečki 1985; 2007). In this chapter we are interested in the intermediary, micro-regional aspect. To be more precise we are interested in the paths that linked the settlement at Tonovcov grad with its surroundings, especially with the main settlements of the time (Kranj, Cividale del Friuli). The study area was selected so that it included all three likely southern and western accesses to Tonovcov grad, i.e. the valley of the river Soča to the south of Kobarid and the two branches of the Nadiža Valley from the spring to the point where it flows into the Friuli plain. In the north the area is limited by the crossroads at which the various paths leading from the Alpine passes meet (Fig. 1.36). As there are no archaeologically documented paths, the historically mentioned paths and the spatial analysis performed with remote sensing and the use of Geographical Information Systems (GIS) were the only sources at the disposal for this study. 1.6.2 HISTORIC SOURCES The oldest relevant data mentioned in historic sources is linked to the so-called Bovec route, a medieval route Sl. 1.36: Tonovcov grad, študijsko območje. Fig. 1.36: Tonovcov grad, the study area. Trbiž. Na to pot so se z južne strani skozi Tolmin in po dolini reke Bače navezovale tudi poti proti osrednji Kranjski, bodisi čez prelaz Sorica v Bohinj, bodisi po dolini Selške Sore proti Škofji loki (Gestrin 1987, 97; glej tam navedeno literaturo). Bovška pot je prvič omenjena v pisnem viru leta 1326. ohranil se je tudi zapis, da je med letoma 1399 in 1404 mesto Čedad začelo graditi cesto čez Predel. Na odseku Trbiž-Bovec so po trasi predhodne tovorniške poti, primerne za ljudi in tovorne živali, zgradili cesto, po kateri so lahko potovali vozovi (Gestrin 1987, 97; Rajšp 1994, 46-47; glej tam navedeno literaturo in vire). Za neposredno okolico Tonovcovega gradu je ta podatek posredno lahko zanimiv. Gradnja ceste namesto tovorniške poti od Bovca dalje bi ne bila smiselna, če v tem času že ne bi bilo ceste mimo Kobarida do Bovca. Ta podatek torej posredno lahko uporabimo kot terminus ante quem za datacijo ceste, ki vodi od Kobarida proti that connected Friuli plain and Carinthia. A rough outline of the route is known: it led from Cividale del Friuli towards the north following the valley of the Nadiža river, through the settlements of Staro selo and Kobarid into the valley of the river Soča to Bovec and further across the Predel/Predil pass and to Treviso. This route was joined by routes that led towards central Carniola via Tolmin and the valley of the river Bača, either across Sorica pass in Bohinj, or along the Selška Sora Valley towards Škofja loka (Gestrin 1987, 97; see cited literature). The first written record of the Bovec route dates back to 1326. Also preserved is the record that states that between 1399 and 1404 the town of Cividale del Friuli started to build a road across Predel. A road suitable for horse drawn carriages was built in the section between Treviso and Bovec, that was previously used only as a track for people and transport animals (Gestrin 1987, 97; Rajšp 1994, 46-47; see cited literature and sources). Sl. 1.37: Tonovcov grad, lidarski posnetek (© ZRC SAZU) prikazan s senčenjem. Puščici kažeta mesti, kjer je bila srednjeveška obsoška pot vsekana v skalo. Fig. 1.37: Tonovcov grad, the lidar data visualization (© ZRC SAZU) shown with hillshading. Arrows are pointing towards the places where the Medieval road was cut into rock. severu neposredno ob Soči in je na dveh mestih vsekana v skalno pobočje (sl. 1.37). Zanimiv je tudi politični kontekst izgradnje t. i. ob-soške ceste, ki so jo vzpostavili v 16. stoletju z namenom neposredne povezave med Gorico in Trbižem. Po več kot pol stoletja načrtovanj so po posredovanju nadvojvode Karla zahtevno gradnjo končali leta 1587. Regionalno je bila izgradnja ceste povezana s povečevanjem prometa čez Gorico na račun Čedada, predvsem pa je s tem habsburška monarhija želela povezati Trst z Notranjo Avstrijo v celoti čez svoje ozemlje, saj je krajša in manj zahtevna Bovška pot deloma potekala po ozemlju Beneške republike (Rajšp 1994, 46-48; glej tam navedeno literaturo in vire). Kontekst dogodkov v 16. stoletju jasno kaže, da je bilo za izgradnjo obsoške ceste od Gorice do Kobarida potrebno posredovanje močne centralne oblasti, ki so jo pri tem vodili strateški interesi. Hkrati je bila neizogibna posledica izgradnje ceste slabitev gospodarskega položaja Čedada. Ali, skozi poznoantično prizmo, Čedad kot močno središče negira obstoj pomembne regionalne povezave This information could be interesting for the case of Tonovcov grad. The construction of the road instead of merely a track from Bovec onwards would not make any sense if the road from Kobarid to Bovec would not have already existed. This data can thus be used as a terminus ante quem for dating the road that leads from Kobarid towards the north along the river Soča and that cuts into the rocky slopes at two points (Fig. 1.37). Also interesting is the political context behind the 16th century construction of the so-called Soča bypass road. This ran along the valley of the river Soča from Gorica to Kobarid from where it continued along the previously described Bovec path across Bovec and Predel all the way to Treviso. Following more than half a century of planning and the intervention by Archduke Charles, the demanding construction was completed in 1587. The road was a result of the desire of the Habsburg monarchy to link Trieste and Inner Austria across its own territory, for a part of the shorter and less demanding Bovec path ran across territories belonging to the Venetian Republic. As a result of this na trasi Gorica-Kobarid. Nasprotno velja za povezavo Kranj-Škofja Loka-Tolmin-Kobarid, ki je bila nujna ob soobstoju Kranja in Čedada kot centralnih krajev. 1.6.3 ANALIZA OPTIMALNIH POTI v arheologiji je iskanje idealnih poti z GIS-i uveljavljena metoda (za pregled glej npr. Connolly, Lake 2006, 252-256). Za analizo morajo biti izpolnjeni vsaj trije pogoji: dovolj kakovostne kartografske podlage (predvsem relief), znana izhodišče in cilj ter znano transportno sredstvo. Prednost uporabe te metode v arheologiji je, da analizo na dovolj razgibanem reliefu lahko opravimo brez kakršnih koli arheoloških podatkov. Šibkost pa je v natančnosti. Za našo analizo to pomeni, da rezultat lahko razumemo kot nekaj deset metrov širok koridor in ne kot dejansko traso poti. Kakovost kartografskih podlag je za analizo optimalnih poti ključen dejavnik (Podobnikar 2009). V primeru pričujoče analize smo razpolagali s podatki nadpovprečne kakovosti. Za osrednji del izbranega študijskega območja smo imeli na voljo izjemno natančen digitalni model reliefa z osnovno celico 0,5 metra in absolutno višinsko in položajno natančnostjo med 0,1 in 0,2 metra (dalje DMR 0,5). Ta je bil izdelan iz podatkov, zajetih z namenskim zračnim laserskim skeniranjem (ang. Light Detection And Ranging; lidar), z algoritmom REIN (Kobler et al. 2007; Kokalj, Oštir, Zakšek 2008). Za celotno študijsko območje smo uporabljali kakovosten digitalni model višin z osnovno celico 12,5 metra (DMV 12,5; Podobnikar 2003). Kot podlage smo uporabljali tudi historične zemljevide, Jožefinske vojaške zemljevide (Rajšp 1997, sekcije 132, 133, 154) in zemljevide Franciscejskega katastra (katastrske občine Kobarid, Staro selo, Ladra, Drežnica in Trnovo). Prvi so bili v merilu 1:28800 izdelani v letih 1763-1787, drugi pa večinoma v drugi četrtini 19. stoletja (arhivirani so dokumenti od 1811 do 1880) v merilu 1:2880. Na podlagi podatkov v pisnem delu omenjenih vojaških zemljevidov iz druge polovice 18. stoletja smo pridobili pomembne podatke o premikanju po pokrajini pred modernimi infrastrukturnimi posegi. Reka Idrija in potoka Učja ter Boka (sl. 1.36) so opisani kot neprehodni, enako Soča na območju soteske med Kobaridom in Trnovim. Območja prodnih plitvin reke Soče južno in severno od soteske so opisana kot prehodna, vendar niso primerna za potovanje vzdolž toka. Ta pogoj smo v analizi simulirali tako, da smo ta območja označili za 20-krat težje prehodna kot kopno z enakim reliefom. Drugače pa smo ravnali v primeru struge Nadiže, po kateri je bilo "ob nizki vodi ^ moč iti večinoma po vodi iz Kreda in Robiča proti Podbeli, po potoku Nadiža pa tudi na Beneško" (transkripcija in prevod vira v Rajšp 1997, 11). road construction the traffic through Gorica increased at the expense of Cividale del Friuli (Rajšp 1994, 46-48; see cited literature and sources). The events in the 16th century clearly show that the construction of the Soča bypass road from Gorica to Kobarid could only take place as a result of the mediation by a strong centralised power that was lead by strategic interests. At the same time the economic weakening of Cividale del Friuli was an unavoidable consequence of the road construction. Or, to look at it through the prism of Late Antiquity, Cividale del Friuli could only exist as a strong centre if there was no important regional link between Gorica and Kobarid. The contrary holds true for the connection Kranj - Škofja Loka - Tolmin - Kobarid, which was a necessity in order for Kranj and Cividale del Friuli to coexist as powerful centres. 1.6.3 ANALYSIS OF OPTIMAL ROUTES In archaeology it is common practice to search for ideal routes with the aid of GIS (for an overview see e.g. Connolly, Lake 2006, 252-256). At least three preconditions have to be fulfilled before an analysis can be performed: the maps need to be of sufficient quality (especially the digital terrain model), the starting point and the final destination of the route have to be known as does the form of transport. The advantage of using this analysis is that it can be performed on a rough relief without any direct archaeological data. The weakness lies in its precision. In our analysis this means that the result can be understood as a corridor a few tens of metres wide rather than the actual route. The quality of the digital terrain model represents a key factor in the analysis of the optimal routes (Podobnikar 2009, 25-29). We had above average quality data, an extremely precise digital terrain model with a basic cell of 0.5 metre and an absolute altitude and positional precision between 0.1 and 0.2 metre (henceforth DT-MDTM 0.5) at our disposal. This was produced from the data obtained from a Light Detection And Ranging (LiDAR) system with an REIN algorithm (Kobler et al. 2007; Kokalj, Oštir, Zakšek 2008). For the entire study area we also used the digital terrain model with a basic cell of 12.5 metres (DTM 12.5; Podobnikar 2003). Historical maps - the first military mapping also known as Joseph II military maps (Rajšp 1997, sections 132, 133, 154) and the maps of the Franciscan cadastre (cadastre municipalities of Kobarid, Staro selo, Ladra, Drežnica and Trnovo) provided an additional source of data. The first were produced between 1763 and 1787 at a scale 1:28800, while the second were mainly produced in the second quarter of the 19th century (the documents were archived between 1811 and 1880) and were mapped at a scale 1:2880. Tok reke Idrije in vseh njenih pritokov je bil v začetku 20. stoletja zaradi melioracije povsem spremenjen. Prvotno strugo smo rekonstruirali na podlagi kartografskega dela Franciscejskega katastra in z vizualno analizo DMR 0,5. Z analizo DMR 0,5, v katerem smo iskali konkavna in konveksna območja, nam je uspelo določiti tudi poti "okrog Sužida (po katerih) je ob mokrem vremenu moč voziti le z lahkimi vozovi", kot jih je opisal podporočnik liškega regimenta Bodoky, kartograf tega območja (transkripcija in prevod vira v Rajšp 1997, 7). Podatke smo analizirali z algoritmom, ki je bil razvit za iskanje optimalnih poti med arheološkimi najdišči ob predpostavki potovanja peš ali s tovornimi živalmi (Podobnikar, Tecco Hvala, Dular 2004). Ta algoritem se je izkazal kot zelo uspešen pri uporabi z DMV 12,5 na podobno razgibanem terenu (Štular 2006b). Poleg DMV 12,5 smo na t. i. karti trenja kot vhodni podatek upoštevali tudi zgoraj opisane dejavnike, ki so omejevali gibanje po pokrajini: neprehodne in prehodne vode, zelo strma skalna pobočja soteske Soče ter območje močvirnih tal. Pri natančnem umeščanju izhodiščnih in končnih točk smo upoštevali načela dobre prakse (Podobnikar 2009). Točke smo umestili na rahlo konkavne mikrolo-kacije, saj v nasprotnem primeru lahko pride do velikih odstopanj v končnem rezultatu zaradi poteka na prvih 50 do 100 metrih. Dodaten ukrep proti tovrstnim napakam in hkrati preverjanje kakovosti smo dosegli tako, da smo vsako pot izračunali v obe smeri. Razhajanja poti "tja in nazaj" so zaradi anizotropnosti algoritma, ki razlikuje med vzponom in spustom, pričakovana (Podobnikar 2009). vendar koherentnost našega rezultata dokazuje, da niti specifičnost algoritma niti izbira izhodiščnih točk nista odločilno vplivali na rezultat. Rezultati se ujemajo s pričakovanji (sl. 1.38). Vse poti po dolini Nadiže potekajo po samem rečnem koritu, enako kot še v 18. stoletju. Od struge Nadiže do Kobarida se pot še pred vasjo Staro selo utiri na stik med ravnino in pobočjem Starijskega vrha. v samem Kobaridu se pot povzpne do zahodnega roba prazgodovinske in rimskodobne naselbine Gradič. Od tam se po manjših dolinah in pobočjih nadaljuje mimo Tonovcovega gradu in se spusti do Soče šele pri Srpenici. Pot prečka Sočo na istem kraju, pri naselju Žaga, kot v 18. stoletju. 1.6.4 MREŽNA ANALIZA Za razliko od analize optimalnih poti so mrežne analize (ang. network analysis) poti v arheologiji redke (npr. Štular 2008; za metodo glej Conolly, Lake 2006, 234-252). Razlog je zelo preprost. Mrežne analize so zasnovane za reševanje problemov, kjer iščemo rešitev med mnogimi možnimi rešitvami. Mrežna analiza, kakršno smo uporabili, med znanimi potmi išče najkrajšo povezavo med izbranim izhodiščem in ciljem ter more- From the written data accompanying the aforementioned military maps we obtained important data as regards the movements through the landscape prior to the modern infrastructure interventions. The river Idrija and the streams Učja and Boka (Fig. 1.36) were described as impassable, as was the river Soča in the gorge between Kobarid and Trnovo. The gravel shallow areas of the Soča river (to the south and north of the gorge) were described as passable but inappropriate for travelling. In our attempt to simulate this condition we marked these areas as 20 times harder to pass than land with the same relief. We applied a different approach in the case of the Nadiža river, where it was 'at low waters... possible to go mainly on water from Kred and Robič towards Podbela, and along the stream all the way to the Venetian lands' (transcription and Slovenian translation of the source in Rajšp 1997, 11). At the beginning of the 20th century the Idrija river bed (as well as all its contributories) was completely changed due to melioration. The course of the original river-bed was reconstructed with the maps from the Franciscan cadastre and a visual DTM 0.5 analysis. With the DTM 0.5 analysis, in which we sought for concave and convex areas, we managed to define the roads 'around Sužid (on which) it was possible to travel only with light wagons at wet weather ', as was described by Bodoky, the cartographer and second lieutenant of the Lika regiment (transcription and Slovenian translation of the source in Rajšp 1997, 7). We analysed the above described data with an algorithm that was developed especially for searching for optimal routes between archaeological sites, assuming that the travel was performed on foot or by transport animals (Podobnikar, Tecco Hvala, Dular 2004). This algorithm has proven itself extremely successful on a similarly diverse terrain when combined with the use of a DTM 12.5 (Štular 2006b). Apart from DTM 12.5 we also took into account (on the so-called friction map) the previously described factors that restricted movement across the landscape: impassable and passable waters, very steep rocky slopes of the Soča gorge and wetlands. Following the precise positioning of the starting and arrival points we took into account the principles of good practice (Podobnikar 2009, 27). The points were placed into slightly concave micro-locations, for otherwise great differences in the final result would occur due to the first 50 or 100 metres of the route. An additional measure performed in order to avoid mistakes (as well as a way to validate the results) was achieved by calculating every route in both directions. Due to the anisotropity of the algorithm, which takes into account the difference between an incline and a decline, the differences in the routes 'there and back' are to be expected (Podobnikar 2009, 27-19). However, the coherence of our result proves that neither the specifics of the algorithm, nor the selection of the starting points had a decisive influence upon the result. Sl. 1.38: Tonovcov grad, rezultati analize optimalnih poti. Fig. 1.38: Tonovcov grad, optimal pathway analysis. bitnimi postanki. V arheoloških raziskavah pa, vsaj kar se tiče gibanja v prostoru, največkrat nimamo nobene vnaprej znane rešitve. Prednost mrežne analize v arheološki raziskavi je natančnost rezultata, saj kot rezultat dobimo natančno traso poti. Slabost pa je, da v analizi ne moremo upoštevati morebitnih neprepoznanih poti. Vendar v primeru tako kvalitetnega vira, kot je DMR 0,5, pričakujemo, da so potencialno neprepoznane le poti, ki so bile trasirane brez posega ali le z minimalnim posegom v relief. Potencialne tovorniške poti pa si na tako razgibanem terenu ne moremo predstavljati brez določenih posegov in vplivov na relief, predvsem pri prečenju pobočij. V primeru okolice Tonovcovega gradu smo z analizo omenjenega DMR 0,5 prepoznali več kot sto različnih poti v skupni dolžini 305 kilometrov (sl. 1.39). Podatek je bil torej kot nalašč za t. i. mrežno analizo. V mrežo poti smo vstavili izhodišča in cilje, tako da smo točkam, The results proved to be in line with expectations (Fig. 1.38). All routes along the Nadiža Valley run along the river bed, which remained the case as late as the 18th century. Between the Nadiža river bed and Kobarid the route sticks to the border between the plain and the slope of Starijski vrh. In Kobarid the route climbs to the western edge of the prehistoric and Roman settlement of Gradič. From there it continues along small valleys and slopes past Tonovcov grad and descends to the Soča river at Srpenica. It crosses the Soča river at the same point, near Žaga, as it did in the 18th century. 1.6.4 NETWORK ANALYSIS In opposition to the analysis of optimal routes, the network analysis of routes is rare in archaeology (e.g. Štular 2008; for the method see Conolly, Lake 2006, Sl. 1.39: Poti, ki so prepoznavne na lidarskem posnetku. Fig. 1.39: Paths recognised on the lidar data visualization. uporabljenim v zgornji analizi, poiskali najbližjo točko na mreži poti. Načrtovali smo večstopenjsko analizo. V prvem koraku smo želeli izvesti enostavno mrežno analizo brez kakršnih koli omejitev. V drugem koraku smo nameravali izločiti poti, ki so bile zaradi strmin neprimerne za tovor-ništvo (za kriterije glej Štular 2006a, 205), in v tretjem koraku še ceste (glej nadaljevanje). V zadnjem koraku smo želeli dodati še postanek pri Tonovcovem gradu. že po izvedenem prvem koraku analize je pot z izhodiščem ob Soči vodila po podobni trasi kot zgoraj opisana optimalna pot. Po drugem koraku, ko smo izločili za tovorništvo neprimerno pot čez Stol, so vse poti vodile po trasah, skoraj identičnih optimalni poti. To pomeni, da je rezultat tudi po tretjem in četrtem koraku ostal nespremenjen, saj rešitev mrežne analize v nobenem koraku ni vodila po cestah, hkrati pa je že vodila mimo Tonovcovega gradu. Slednje pomeni, da se pot ni prilagajala lokaciji Tonovcovega gradu, temveč obratno. Če bi se trasa poti prilagajala lokaciji Tonovcovega gradu, bi idealna pot vodila vstran in bi se približala lokaciji šele v četrtem koraku, ko smo upoštevali obvezen postanek pri naselbini. 234-252). The reason behind this is extremely simple. Network analysis is used to find the shortest connection between the starting and ending point (and the possible stops in between) amongst the known routes. In archaeological research - at least when dealing with movement in space - no solutions are given in advance. In archaeological research the advantage of network analysis lies in the precision of its result, for the result is reflected in an exact route. The disadvantage lies in the fact that we cannot take into account the possible previously unrecognised routes. However, when we have such a high quality source at our disposal as a DTM 0.5, we expect that merely routes that were created without any intervention into the relief could be potentially unrecognised. On such a rough terrain as the one analysed any potential transport route could not be imagined without certain interventions into the relief, especially when crossing the slopes. The analysis of the aforementioned DTM 0.5 in the surroundings of Tonovcov grad has shown over one hundred possible routes in a total length of 305 kilometres (Fig. 1.39). This data is thus perfect for a network analysis. We entered the starting and ending points of 1.6.5 HISTORIČNA ANALIZA Zgoraj opisano mrežo vseh poti smo analizirali tudi s stališča konteksta. Poglavje smo sicer naslovili historična analiza, saj smo najpomembnejše podatke pridobili na podlagi pisnih virov in historičnih zemljevidov. Vendar smo večino poti interpretirali na podlagi konteksta. postopek najlaže opišemo kar na konkretnih primerih (sl. 1.40). Najlaže je bilo izločiti poti, nastale v 1. svetovni vojni, saj vodijo neposredno do položajev, kjer se tudi končajo. Podobno smo na podlagi različnih kriterijev interpretirali tudi večino preostalih poti: Pohodniške poti - imajo izhodišče in/ali cilj na modernih cestah (stratigrafski odnos mlajše od) - sekajo in/ali delno izkoriščajo poti iz 1. svetovne vojne (stratigrafski odnos mlajše od) - pogosto gre za najožje izmed dokumentiranih poti, poseg v pobočje pri prečenju je zanemarljiv - pogost cilj je vrh ali razgledna točka - pogoste so krožne poti - pogoste so številne variante na kratkih odsekih. Gospodarske (vaške) poti - povezujejo naselbine s pripadajočimi gospodarskimi objekti, torej vasi s polji, senožetmi in planinami - gostota mreže teh poti je sorazmerna z intenzivnostjo gospodarskega objekta, tj. najgostejša mreža na poljih, najredkejša v strmih pobočjih s senožetmi - praviloma so vrisane v Franciscejski kataster. Ceste - velik poseg v okolje in posledično velik vložek dela (useki v skalna pobočja, poti s številnimi mostovi) - (lahko prepoznavnih modernih cest v analizo nismo vključili). Stare poti - smo poimenovali poti, ki ne sodijo v nobeno izmed zgornjih kategorij. Sklepamo namreč, da smo z zgoraj opisano erudicijo vira izluščili vse poti razen najstarejših na srednje ali dolge razdalje. Vendar je ta korak v določeni meri tudi subjektiven. Najpogosteje so problem gospodarske poti, ki se povezujejo v poti na dolge razdalje. Tak primer so na primer gospodarske poti Kobarida in Idrskega. Na odseku poti severno od Tonovcovega gradu smo v analizo skušali pritegniti tudi topografski podatek o rimski cesti oziroma ledinsko ime Pod staro potjo južno od naselja Trnovo ob Soči (sl. 1.40). Edina pot v bližini je moderna cesta, ki na tem delu poteka po ne- the routes into the network by pinpointing the points used in the previous optimal path analysis. We planned a multilevel analysis. In the first step a simple network analysis without any restrictions was performed. In the second step the routes that were inappropriate for animal transport due to the steep slopes (for criteria see Štular 2006a, 205) were excluded and in the third step the roads were excluded (see below). In the final step the stop at Tonovcov grad was added. Following the first step the route starting alongside the Soča led along a similar route as the previously described optimal path. After the second step (in which we eliminated the route across Stol as it was inappropriate for transport) all routes led along routes that were almost identical to the optimal path. This means that the result remained unchanged also after the third and fourth step, for the solution provided by the network analysis did not lead along the road at any stage, and it already lead past Tonovcov grad. In other words, steps three and four were not necessary. The interpretation of the latter leads us to the conclusion that the path did not adjust to the location of Tonovcov grad, but the other way round. If the route were adjusted to pass Tonovcov grad, the ideal route would lead to the side and would come close to the location only in stage four, when we took into account the obligatory stop at the settlement. As the results show, the location of Tonovcov grad has been chosen in the proximity to the optimal route. 1.6.5 HISTORICAL ANALYSIS The above described network of all routes was also analysed from the contextual aspect. This chapter was entitled Historical analysis, for the most important data was obtained from written sources and the contextual interpretation of historical maps. The context in question has been obtained from DTM 0.5. The procedure can be best described as a regression analysis and is best explained on actual examples (Fig. 1.40). The easiest step was to remove all routes that were created during World War I, for they led to military trenches at which they ended abruptly. Most of the remaining routes were similarly interpreted on the basis of various criteria: Hiking paths - Have a starting and/or ending point on modern roads (stratigraphic relationship younger than); - They cross or partially make use of the World War l routes (stratigraphic relationship younger than); - Often they are the narrowest of the documented routes, and they create a minimum impact when crossing a slope; ceste / roads pohodniške poti / hiking paths poti zgrajene v 1. svetovni vojni / World War I paths gospodarske poti / economic paths stare poti / old tracks Sl. 1.40: Poti, retrogradna analiza. Fig. 1.40: Paths, retrograde analysis. spremenjeni (vsaj) poznosrednjeveški trasi. Pridevnik "stara" oziroma v varianti "rimska" se torej nanaša na predhodnico te poti. Poleg omenjene ceste so na tem območju le poti, ki jih je podporočnik Bodoky (transkripcija in prevod vira v Rajšp 1997, 7) opisal kot "strme ozke poti, ki vodijo v senožeti". Slednje imajo tudi priključek na pobočno pot. Opisan položaj lahko razložimo le na dva načina. Verjetneje je, da podatek kaže na lokalni priključek na staro pot, saj gre za povsem nelogičen ovinek. Druga možna je razlaga, da predpona pod označuje odnos ledine pod staro potjo. - The path often ends on top of a peak or at a vista point; - Circular paths are common; - Numerous options can be found on the steep sections. Economic (settlement) paths - Link settlements with the economic objects belonging to it, i.e. villages with fields, meadows and pastures; - The network density of these paths is in accordance to the intensity of the economic object, i.e. it is the densest in the fields and the least dense in the steep slopes with meadows and pastures. - The majority of them have been mapped for the Franciscan cadastre. 1.6.6 VREDNOTENJE Rezultati analiz so si zelo podobni. Ko jih prikažemo na istem zemljevidu (sl. 1.41), vidimo, da v neposredni bližini Tonovcovega gradu prikazujejo isti koridor. S kombinirano uporabo analize optimalnih poti in mrežne analize smo dosegli združevanje prednosti in izničenje slabosti obeh metod. Roads - Show great interventions into the environment, are the product of large investment of labour (cuts into the rocky slopes, routes with numerous bridges); - (the easily recognisable modern roads were not included in the analysis). Sl. 1.41: Tonovcov grad, rezultat analize optimalnih poti, mrežne analize in vidljivosti. Fig. 1.41: Tonovcov grad, the results of optimal pathway analysis, network analysis and visibility. Podobno lahko ugotovimo za celoten odsek razgibanega reliefa z minimalnimi modernimi infrastruk-turnimi posegi od Kobarida mimo tonovcovega gradu do Srpenice. Prikaz rezultata analize optimalne poti na podlagi DMR 0,5 namreč jasno pokaže, da so odstopanja posledica nenatančnosti DMV 12,5. tovorniška pot je torej vodila po trasi, ki jo kaže rezultat mrežne analize. Natančnost zagotavlja mrežna analiza, optimalnost trase ne glede na danes ohranjene poti pa analiza optimalnih poti. Samo dejstvo, da smo do istega rezultata prišli s tremi neodvisnimi metodami in dvema neodvisnima viroma, potrjuje pravilnost rezultata. Predvsem možnost, da bi bilo ujemanje prve analize z drugima dvema naključno, je zaradi uporabe različnih virov (DMR 0,5 in DMV 12,5) skorajda nična. Natančneje, možnost naključnosti je 0,2 %1. Nenaključnost rezultata sama po sebi seveda ne priča o starosti poti, temveč zgolj o primernosti te poti 1 Verjetnost naključnega ujemanja dveh binarnih rastrov v eni celici je 50 % (kombinacije so 0 0, 0 1, 1 0, 1 1). Verjetnost naključnega ujemanja 9 celic je torej 1/29, kar je približno 0,2 %. Old routes - The routes that did not belong into any of the above categories were considered old routes. We concluded that with the above described erudition of the source we eliminated all routes except for the oldest middle and long distance ones. However, to a certain extent this step is subjective. The most common problem is represented by the economic routes that partially overlap with the long distance routes. Such an example is for instance represented by the economic routes in the vicinity of Kobarid and Idrsko. On the section of the route north of Tonovcov grad we tried to include the data stemming from archaeological topography on the place name Roman road or the old Slav name Pod staro potjo (Under the old path), south of the settlement of Trnovo ob Soči (Fig. 1.40). The only existing route in the vicinity of this two place name path is the modern road that has in this section not changed since the Late Medieval times. The adjective 'old' or in one variant 'Roman' thus denominates a predecessor to this route. Apart from the aforementioned road only tracks described by second lieutenant Bodoky as 'steep narrow paths that lead to meadows' (transcription and Slovenian Sl. 1.42: Prerez poti med Kobaridom in Srpenico, primerjava stare poti (zgoraj) in v srednjem veku zgrajene ceste (spodaj). Fig. 1.42: Profile of the path between Kobarid and Srpenica; comparison of the old track (above) with the road built in the Middle Ages (bellow). za potrebe tovorniškega prometa. Gre le za statistično potrditev, da na območjih s tako razgibanim reliefom možnih rešitev ni veliko. Zato tudi pravilno ovrednotene srednjeveške vire lahko uporabljamo kot analogijo za starejša obdobja. o starosti poti pa lahko sklepamo na podlagi dejstva, da te poti povezujejo železnodobne naselbine Tolmin, Kobarid in Bovec s sočasnima naselbinama ob Nadiži - (verjetno) Sveti Volar nad Robičem in Špeter (sl. 1.36; prim. Mlinar 2004; Mlinar, Pettarin 2007). Pri tem velja poudariti, da lokacije slednjih pri analizah nismo upoštevali in gre torej za še en neodvisen podatek, ki potrjuje pravilnost trase. S stališča najdišča Tonovcov grad je najpomembnejši rezultat te analize določitev trase poti med Kobaridom in Srpenico pred izgradnjo ceste, ki je mestoma vklesana v skalna pobočja soteske Soče. Stara pot je na tem odseku krajša za 1,39 kilometra oziroma 15 %, vendar premaga 390 metrov višinske razlike namesto 180 (sl. 1.42). Slednje jasno kaže na prednosti trase ceste. Po vsem sodeč je bil torej nastanek naselbine na Tonovcovem gradu povezan tudi s traso tovorniške poti, ki je povezovala Čedad čez Nadižo, Bovec in alpske prelaze s Koroško. Je pa ta naselbina v primerjavi s starejšo lokacijo Gradič nad Kobaridom umaknjena od poti, ki je povezovala Kranj s Čedadom čez Bačo, Tolmin, Kobarid in Nadižo. Kljub temu so imeli prebivalci Tonovcovega gradu pregled vsaj nad ključnim delom te poti, prehodom čez Sočo. Prehod zgornje analize sicer postavljajo nekoliko južneje od Kobarida, a glede na rezultate avtorjevih terenskih ogledov se zdi lokacija ob samem vznožju današnjega Kobarida najverjetnejša. translation of the source in Rajšp 1997, 7) can be found in this part. These paths lead to the ridge-path that has been identified as the optimal path in the previous analyses. The described situation can be interpreted in two ways. It is more likely that the place names describe the local path leading to the old route. The actual digression of the old route from the ridge path seems rather unlikely. The other possible interpretation is that the prefix 'under' denotes the relation of the place name location as lying directly under the old path. This would be a good description of the actual topographic position if the place name location is situated in the valley under the ridge path that has been identified as the optimal path within the previous analyses. 1.6.6 EVALUATION The various analyses yield very similar results. when depicted on the same map (Fig. 1.41) we can see that they run along the same corridor in the direct vicinity of Tonovcov grad. With a combined use of the analysis of optimal routes and the network analysis we merged the advantages and annulled the disadvantages of the two methods. Similar can be ascertained for the entire section from Kobarid past Tonovcov grad to Srpenica, which has a diverse relief and shows minimal modern infra-structural interventions. The depiction of the result obtained by the analysis of the optimum route on the basis of DTM 0.5 clearly shows that the deviations of the optimal path calculated on the basis of DTM 12.5 are indeed a consequence of the DTM 12.5 inaccuracies. The track used by transport animals (but not wagons) thus ran along the route as shown in the network analysis result. The precision of the result is ensured by the network analysis, while the validity irrespective of the modern preservation is ensured by the optimal routes analysis. The result is verified by the fact that we achieved the same result with the use of three independent methods and two independent sources. It is almost impossible for the first analysis to yield the same results as the second. To be precise the possibility for this being a coincidence is 0.2 %1. Of course, the non-randomness of the result does not indicate the age of the route, but merely indicates the appropriateness of this route for the needs of transport. It is merely a statistical confirmation that there are only a few possible solutions in areas with such a diversified 1 The possibility of a coincidental match of two binary rasters in a single cell is 50 % (combinations are 0 0, 0 1, 1 0, 1 1). The results of the optimal path analysis and network analysis match in 9 cells. The possibility of a coincidental match of nine cells would therefore be H9, which is roughly 0.2 %. Analiza vidnosti potrjuje pregled nad prehodom reke in hkrati razkriva, da s Tonovcovega gradu ni možno opazovati poti Kobarid-Nadiža v smeri iz Čedada ali proti njemu (sl. 1.41). relief. Thus, the correctly evaluated medieval sources can be used as an analogy for the older periods. We can ascertain the age of the path on the basis of the fact that these paths connected the Iron Age settlements of Tolmin, Kobarid and Bovec with the settlements alongside the river Natisone, i.e. (probably) Sveti Volar nad Robičem and San Pietro al Natisone [Slovenian Špeter] (Fig. 1.36 ; cf. Mlinar 2004; Mlinar, Pettarin 2007). At this it should be emphasised that the locations of the Iron age hillforts were not taken into account in the analysis and that this therefore represents additional independent verification for the validity of the route. From the aspect of the site at Tonovcov grad the most important result of the analysis lay in the definition of the route between Kobarid and Srpenica prior to the road (in some places carved into the rocky slopes of the Soča gorge) construction. In this section the path that predated the medieval road was shorter by 1.39 kilometres (15 %), however it has to overcome 390 altitude metres instead ofthe 180 covered by the road (Fig. 1.42). The latter clearly shows the advantages of the road, i.e. it is passable by wagons. Taking everything into account the origin of the settlement at Tonovcov grad was obviously also linked to its position adjacent to the transport route that linked Cividale del Friuli with Carinthia (across the Nadiža, Bovec and the Alpine passes). When compared to the older settlement location Gradič nad Kobaridom, Tonovcov grad was removed from the route that linked Kranj with Cividale del Friuli (across Bača, Tolmin, Kobarid and the Nadiža). Regardless of this the inhabitants of Tonovcov grad had a clear view of one of the key sections of the route, i.e. the crossing of the Soča river at Kobarid. The above analyses are showing the river crossing somewhat south of the present day Kobarid, but the author's ground truthing suggests the location just beneath the Kobarid. This is revealed by the visibility analysis (Fig. 1.41), which also revealed that it was impossible to see the route Kobarid - the Nadiža (in the direction to or from Cividale del Friuli) from Tonovcov grad. 2. terenski izvid 2. field report 2.1 Metodologija dela 2.2 Kronologija in faze 2.3 stavba 1 z okolico 2.4 stavbi 2 in 3 2.5 sklop cerkva 2.6 Cisterna (vodni zbiralnik) 2.1 Methodology 2.2 Chronology and phases 2.3. Building 1 and its surroundings 2.4 Buildings 2 and 3 2.5 The ecclesiastical complex 2.6 Water cistern (reservior) GK koordinate / coordinates E 390 917,26 N 124 676,73 ETRS89/TM koordinate / coordinates E 390 546,22 N 125 164,94 Sl. 2.1: Načrt najdišča. M = 1:1000. Fig. 2.1: Map of the site. scale = 1: 1000. 2.1 metodologija dela 2.1 methodology pred začetkom izkopavanja je bila na Tonovcovem gradu narejena geodetska izmera. postavljenih je bilo več fiksnih geodetskih točk, ki so določale koordinatni sistem z linijama y (smer sever-jug) in x (smer vzhod-zahod). Na osnovne linije je bila potem postavljena mreža kvadrantov s stranicami 4 x 4 m (sl. 2.1), znotraj izkopnih polj pa so bili kvadranti razdeljeni še na mikrokvadrante s stranicami 1 x 1 m. višine so bile do leta 2002 merjene relativno od dveh absolutno izmerjenih geodetskih točk (399,55 m n. m. na območju izkopnega polja stavbe 1 ter 410,25 m n. m. na območju cerkvenega sklopa), po letu 2002 pa absolutno. Tudi relativno merjene višine so v knjigi preračunane v absolutne nadmorske višine. vrh Tonovcovega gradu s poznoantično naselbino je zelo neraven. v osnovi se deli na tri dele. Najnižji je severozahodni del naselbine. Teren se proti jugu nato strmo dvigne za približno 10 m in se zravna v raven, izpostavljen plato, na katerem je bil postavljen sklop cerkva, na vzhodnem delu pa se tik nad sočo dviga najvišji plato (412 m. n. m, glej tudi pogl. 1.1 in 1.3). na strmih pobočjih je bil humusni pokrov zelo tanek, tako da je bila na nekaterih mestih že pred arheološkimi posegi vidna skalna osnova, medtem ko so bile drugje med izkopavanji odkrite tudi več metrov globoke skalne kotanje. nekatere so bile zapolnjene s sterilnimi plastmi gline in gruščev, druge pa so vsebovale bogate kulturne plasti. Hrib je bil pred začetkom raziskav prekrit z redkim gozdom (bukev, gaber, hrast) in grmičevjem. plast humusa, ki je prekrivala ruševine poznoantičnih stavb, je bila zelo tanka, na nekaterih mestih pa je sploh ni bilo. osnovne oblike zidanih objektov so bile zato ponekod na površini še jasno razvidne, tako da se je v grobem dalo določiti njihove obrise (glej pogl. 1.3, sl. 1.6, 1.9, 1.10, 1.12). na podlagi tako vidnih objektov so bila potem izbrana območja raziskav. zaradi strmega terena so bile posebno v ruševin-skih plasteh najdbe pomešane, saj so predmeti z višje ležečih predelov padli ali spolzeli na nižje ležeče. Tako lahko za veliko število novcev in nekaj razbitih tegul na sicer neposeljenem pobočju nad stavbo 1 domnevamo, da so prišli sem z višje ležečega cerkvenega platoja. prior to the beginning of the excavations a land survey was conducted at Tonovcov grad. A number of fixed geodetic points were positioned and they were used to define a coordinate system with the y (north - south direction) and x (east - west direction) axes. A grid of quadrants measuring 4 x 4 metres was placed along the basic lines (Fig. 2.1), and within the excavation area the quadrants were divided into micro quadrants measuring 1 x 1 m. Until 2002 all heights were measured relatively by using two previously absolutely measured geodetic points (399.55 m a.s.l. in the excavation area of building 1 and 410.25 m a.s.l. in the area of the ecclesiastical complex), while after 2002 all heights were measured absolutely. For the purpose of this book all relatively measured altitudes were recalculated into absolute altitudes. The peak of Tonovcov grad with the Late Antique settlement on it is very uneven. The top of the hill can be divided into three parts, the lowest of which was occupied by the northwest part of the settlement. From here the terrain rises sharply towards the south (by approximately 10 m) and levels out into a straight and exposed plateau, the top of which was covered by ecclesiastical complex. The highest plateau on the east rises above the soča river (412 m.a.s.l., see also chapters 1.1 and 1.3). The steep slopes were covered by a very thin layer of humus, in some spots revealing the rock base already prior to the archaeological excavations, while elsewhere up to a few meters deep sinkholes were discovered during the excavations. some of them were filled with sterile layers of clay and gravel, while others included rich cultural layers. prior to the beginning of the research the hill was covered in trees (beech, hornbeam and oak trees) and bushes. The humus layer that covered the ruins of the Late Antique buildings was very thin or even nonexistent. Thus the basic forms of some of the structures were visible on the surface and their outlines could be at least roughly defined (see chapter 1.3, Figs. 1.6, 1.9, 1.10,1.12). The research areas were selected on the basis of the visible structures. Due to the steepness of the terrain the finds were very mixed, especially in the destruction layers, for Korenine dreves in grmovja, ki so preraščali ruševine objektov, so bile zelo razvejene in so zato na nekaterih mestih močno poškodovale arheološke plasti. Pri odstranjevanju koreninskih sistemov je bilo velikokrat poleg same ruševine poškodovanih tudi nekaj plasti pod njo. problem je predstavljala tudi ločljivost plasti. posebno zgornje, humusno-ruševinske, so si bile po barvi in strukturi zelo podobne, razlikovanje med njimi pa so oteževale tudi poškodbe zaradi korenin in kasnejših posegov. Ti so bili dokumentirani v stavbi 1, na več mestih v cerkvenem sklopu in v vodnem zbiralniku (glej pogl. 2.3, 2.5, 2.6). Metodologija izkopavanj se je zato morala prilagoditi danemu stanju. Kjer so bile plasti nepoškodovane in jasno razvidne, je izkopavanje potekalo po plasteh. Najdbe znotraj plasti so bile dokumentirane tudi v okviru mreže mikrokvadrantov z metrsko natančnostjo, posebne najdbe pa so dobile še koordinate x, y in z. Najdbe iz nejasnih ali premešanih plasti pa so bile prostorsko umeščene samo znotraj mreže mikrokva-drantov, brez določitve plasti, dokumentirana je bila tudi višina. Tudi v teh primerih so posebne najdbe dobile še koordinate x, y in z. pri izkopavanjih je bil kot pomoč pri iskanju kovinskih najdb uporabljen detektor kovin. plasti so bile pregledane z detektorjem že pred odstranjevanjem, vzorec pa je bil nato, ko je bil že odstranjen, pregledan še enkrat. Ta način se je izkazal kot najuporabnejši pri zgornjih, ruševinskih plasteh, v katerih so prevladovali kamni. Sejanje ali spiranje se je pri takih vzorcih izkazalo za neekonomično. pri plasteh, kjer je njihova sestava to omogočala, pa je bilo uporabljeno tudi suho sejanje.1 plasti v izkopnih poljih so bile v večini primerov raziskane do geološke osnove. Izjema je bilo samo izkopavanje cerkvenega sklopa, kjer je bila zaradi odlično ohranjenih estrihov v severni in osrednji cerkvi sprejeta odločitev, da se v čim večji meri ohranijo in zaščitijo (glej pogl. 2.5). Globino in sestavo plasti pod estrihi smo lahko tako raziskali le z manjšimi kontrolnimi sondami v ladjah severne in osrednje cerkve. Terenska dokumentacija je obsegala terenske dnevnike, opise plasti, sezname najdb, risbe planumov in presekov ter fotodokumentacijo. the objects from the top of the slope fell or slid to the lower lying parts. Thus it can be assumed that the large number of coins and some of the broken tegulae on the slope over building 1 where there are no structures fell from the church plateau. The roots of the trees and bushes that grew over the ruins spread out and caused severe damage to the archaeological layers in some spots. As the roots were removed it was often revealed that not only the ruins but also several layers underneath were destroyed. The distinction between various layers represented an additional problem. Especially the upper destruction layers proved to be very similar in colour and texture, and distinguishing between them was made harder by the damage created by the roots and the later interventions. These were documented in building 1, on a number of places within the ecclesiastical complex and alongside the water reservoir (see chapters 2.3, 2.5, 2.6). The methodology of the excavations thus had to be adjusted accordingly. Where the layers were undamaged and clearly visible the excavation proceeded by layers. The finds in a layer were documented within a microquadrant grid with metre preciseness, and special finds obtained coordinates x, y and z. If the layers could not be precisely defined the finds were recorded merely within the micro-quadrant grid, and their z-coordinate was documented. The special finds again obtained their x, y and z coordinates. A metal detector was used for collecting metal finds. The layers were checked with a metal detector prior to their removal, and the sample was checked once again after it was removed. This has proven to be useful in the upper, destruction layers, in which stones dominate. Sieving and flotation have proven to be inefficient with such samples. Dry sieving was used wherever the composition of the layer permitted.1 In most cases the layers in the excavation area were researched all the way to the geologic base. An exception to this rule was the excavation of the ecclesiastical complex in which the mortar floor in the north and main church was excellently preserved, thus a decision was reached that the floors are to be preserved and protected as well as possible (see chapter 2.5). The depth and composition of the layer under the mortar floor were thus defined on the basis of smaller control trenches in the naves of the north and central church. The documentation consisted of field diaries, layer descriptions, lists of finds, drawings of planums and sections as well as photo documentation. 1 To so predvsem spodnje plasti na izkopnem polju stavbe 1, medtem ko pri izkopavanju cerkvenega sklopa in objektov 2 in 3 ta način ni bil uporabljen. ' These are mainly the lower layers on the excavation area of building 1, while at the excavation of the ecclesiastical complex and buildings 2 and 3 this method was not used. 2.2 kronologija in faze 2.2 chronology and fhases Na podlagi dosedanjih raziskav je bilo mogoče na Tonovcovem gradu določiti pet obdobij poselitve (prazgodovina, antika, pozna antika, zgodnji srednji vek, srednji vek). nekatere izmed njih je bilo mogoče razdeliti tudi na faze (predvsem velja to za najbogatejše poznoantično obdobje). pri tem je treba poudariti, da nekatera obdobja za zdaj nimajo jasne potrditve tudi v sami stratigrafski situaciji na najdišču. Tako prazgodovinsko (železnodobno), antično in srednjeveško poselitev predvidevamo predvsem na podlagi posamičnih najdb iz teh obdobij, najdenih v plasteh obeh poznoantičnih faz, ter še neraziskanih ostankov arhitekture (domnevni srednjeveški stolp v zahodnem delu naselbine). pri časovni določitvi obdobij in faz so bile uporabljene samo najdbe, ki izvirajo iz zanesljivih plasti. vse ostale najdbe, kot so npr. detektorske, pobrane pred začetkom izkopavanj zunaj izkopnih polj (glej pogl. 1.4), in tiste iz poškodovanih ali nezanesljivo razbranih plasti, so bile sicer obdelane v okviru analize gradiva in kot pomoč pri interpretaciji, niso pa bile uporabljene pri datacijah faz, kartah razprostranjenosti predmetov po objektih in statističnih analizah. 2.2.1 prazgodovina nad plastmi geološke osnove (skalna osnova, skalna preperina, sterilne gline) so bile na več mestih dokumentirane plasti, ki so vsebovale skromne prazgodovinske ostanke. Take plasti so bile na območju stavb 1, 2, 3 in cisterne. V vseh primerih gre za plasti trde, oran-žno-rdeče gline, ki je bila zelo težko ločljiva od sterilne glinene osnove. najdbe iz plasti predstavljajo sileksi, nekaj kamnitih orodij in nekaj zelo fragmentirane in slabo ohranjene prazgodovinske keramike (pogl. 2.3.1, 2.4.1; glej tudi Tonovcov grad. Najdbe, pogl. 6). Že na vrhu prazgodovinskih plasti pa so se začenjale pojavljati tudi antične najdbe (včasih celo neposredno skupaj). Poleg tega so bila nekatera kamnita orodja najdena tudi kot rezidualne najdbe v mlajših, antičnih plasteh. Ta prazgodovinska poselitev je datirana v dolgo časovno obdobje med paleolitikom in bronasto dobo. znotraj The research carried out at Tonovcov grad has revealed five settlement periods: prehistory, Antiquity, Late antiquity, Early Middle ages and Middle ages. some of them could be divided into various phases (especially the late antique period, which revealed the most finds). Here it should be stressed that some periods have so far not been confirmed in the stratigraphic situation at the site. Thus the prehistoric (Iron age), antique and high Medieval settlement was assumed on the basis of individual finds belonging to these periods, which were found in the layers of the two late antique phases and the as yet uninvestigated architectural remains (the assumed Medieval tower in the west of the settlement). only finds that originate from reliable layers were used to define the time scales of the various periods and phases. A11 other finds - such as the finds discovered with the use of a metal detector that were gathered prior to the excavations and outside the excavation areas (see chapter 1.4) or finds from damaged or unreliably dated layers - were treated within the material analysis and provided help in the interpretation, however they were not used to date the phases, nor were they shown in the distribution maps of finds or in statistical analyses. 2.2.1 prehistory The layers above the geological base (bedrock, decayed bedrock, sterile clays) included modest prehistoric remains. Such layers were found in the areas of buildings 1, 2, 3 and the water cistern. In all cases these were layers of compact orange-red clay that were very hard to separate from the sterile clay base. The finds from these layers are represented by a combination of silex, stone tools and extremely fragmented and poorly preserved prehistoric pottery (chapters 2.3.1, 2.4.1; see also Tonovcov grad. Finds, chapter 6). antique finds appear already on the top of these prehistoric layers, sometimes together with the prehistoric ones. Stone tools also appear as residual finds in the later antique layers. This prehistoric settlement is dated into the long period between the Paleolithic and Bronze age. There tega časa obstaja več poselitvenih faz, ki pa jih zaradi slabe ohranjenosti plasti in najdb za sedaj ne moremo natančneje opredeliti. Verjetna je poselitev v mezolitiku in bronasti dobi (glej Tonovcov grad. Najdbe, pogl. 6.5). Najdeno je bilo tudi nekaj predmetov, ki so datirani v starejšo in mlajšo železno dobo (glej Tonovcov grad. Najdbe, pogl. 6.2, 6.3; sl. 6.2). To so dokaj reprezentativne najdbe, ki pa so bile vse odkrite v mlajših, že poznoantičnih plasteh. Zanesljivih poselitvenih plasti iz železne dobe na do sedaj raziskanih območjih nismo odkrili. 2.2.2 ANTIKA Poselitev iz časa 1. in 2. st. ni zanesljivo potrjena z naselbinskimi ostanki. v ta čas bi po stratigrafskih odnosih sicer lahko sodile najstarejše antične plasti pod stavbo 3 (SE 177, 179, glej pogl. 2.4.1, 3.2.2), vendar je potrebna previdnost pri opredeljevanju zgodnjih antičnih najdb, posebno fibul, ki se na Tonovcovem gradu pogosto pojavljajo v mlajših plasteh. v poznoantičnih (predvsem ruševinskih) plasteh stavb 1 in 3 je bilo najdenih nekaj novcev in fibul iz tega obdobja. Prva poselitvena faza višinskih naselbin v Sloveniji, čas zadnje tretjine 3. stoletja (Ciglenečki 1990, 154-156; 1999, 292), je na Tonovcovem gradu zastopana predvsem s posamičnimi najdbami. Edine plasti, ki bi lahko sodile v ta čas, so bile ohranjene pod stavbo 3 (SE 172, 175). V SE 172 je bil najden novec, datiran v leto 272 (glej pogl. 2.4.1 in Tonovcov grad. Najdbe, pogl. 5.1, kat. št. 22). Ostalo dobro opredeljivo gradivo tega časa je bilo najdeno v plasteh prve (večina novcev) in druge poznoantične faze na območju stavbe 1 (obročaste fibule, dva ključa v obliki prstana, pasna spona ter ščitnik noža; Tonovcov grad. Najdbe, pogl. 2.1). 2.2.3 POZNA ANTIKA PRVA POZNOANTIČNA FAZA (PA 1) čas pozne antike pomeni vrhunec poselitve To-novcovega gradu. Razdelimo ga lahko na dve fazi, ki sta obe jasno vidni tudi iz stratigrafskih odnosov. Starejša, prva poznoantična faza (PA 1) sodi v čas druge polovice 4. in v začetek 5. st., njen konec pa časovno ni povsem dobro omejen. To obdobje (imenovano tudi poznorim-sko; o problematičnosti terminologije glej Ciglenečki 1999, 290-292) je v jugovzhodnih Alpah čas nastanka številnih utrjenih vojaških postojank, katerih namen je bil varovanje strateško pomembnega območja ob vstopu v Italijo (Ciglenečki 1999, 291-292). Na Tonovcovem gradu so bili ostanki iz tega časa najdeni pod stavbama 1 in 2. Faza je dobro dokumentirana tudi s keramičnimi, kovinskimi in novčnimi najdbami iz mlajših plasti. Na were a number of settlement phases within this period, however, due to the poor preservation of the layers and finds it is impossible to establish this more precisely. Settlement in the Mesolitic period and in the Bronze Age is probable (see Tonovcov grad. Finds, chapter 6.5). A few objects that could be placed into the early or late Iron Age were also discovered. These are relatively representative finds that were all discovered in the later, Late Antique layers (see also Tonovcov grad. Finds, chapters 6.2, 6.3; Fig. 6.2). In the areas researched so far we have not found any reliable settlement layers from this period. 2.2.2 ANTIQUITY The discovered remains are not enough to reliably confirm settlement on the hill in the 1st and 2nd centuries. Due to their stratigraphic relations the oldest Antique layers under building 3 (SU 177, 179, see chapters 2.4.1, 3.2.2) could be dated into this period, however great care is needed when defining layers on the basis of early Antique finds, especially fibulae, which represent a common find in the younger layers at Tonovcov grad. Finds from this period were also discovered in the Late Antique (especially destruction) layers of buildings 1 and 3. At Tonovcov grad the first settlement phase of the hilltop settlements in Slovenia, i.e. the last third of the 3rd century (Ciglenečki 1990, 154-156; 1999, 292), is also represented mostly by individual finds. The only layers that could belong into this period were preserved under building 3 (SU 172, 175). A coin dated to 272 AD (see chapter 2.4.1 and Tonovcov grad. Finds, chapter 5.1, Cat. No. 22) was found in SU 172. The remaining material that can be reliably dated to this period was found in the layers of the first (mainly coins) and second (ring fibulae, two rings-keys, a belt buckle and a knife hilt plate) Late Antique phases in the area of building 1 (Tonovcov grad. Finds, chapter 2.1). 2.2.3 LATE ANTIQUITY LATE ANTIQUITY PHASE 1 (LA 1) The Late Antique period represents the pinnacle of the settlement at Tonovcov grad. It can be divided into two phases, both of which can be clearly seen in the stratigraphy. The earlier Late Antiquity phase (LA 1) is dated to the second half of the 4th and beginning of the 5th century. The end of this phase could not be clearly defined. This period (also known as Late Roman; for the terminology issues see Ciglenečki 1999, 290-292) was a period in which numerous fortified military posts appeared in the Southeastern Alps in order to guard the strategically important entrance into Italy (Ciglenečki območju cerkvenega sklopa ter vodnega zbiralnika plasti te faze niso bile najdene. DRUGA POZNOANTIČNA FAZA (PA 2) Mlajša, druga poznoantična faza (PA 2) je še bogatejša in bolje ohranjena od prve. Trajala je predvidoma med koncem 5. in začetkom 7. stoletja. V to obdobje lahko umestimo večino raziskanih objektov v naselbini (stavbi 1 in 2, cerkveni sklop, vodni zbiralnik) ter veliko večino kovinskega in keramičnega drobnega gradiva. 2.2.4 ZGODNJI SREDNJI VEK (ZSV) Kot trenutno še ne dobro razumljiva se kaže poselitvena faza, ki jo lahko grobo datiramo v zgodnjesrednjeveško obdobje. Vanjo lahko uvrstimo nekatere drobne najdbe, nekaj plasti v stavbi 1, kurišče s keramiko v glavni cerkvi, grob 21 iz cerkvenega sklopa, skupino grobov ali vsaj enega od njih (grob 18) za južnim zidom stavbe 2 ter kurišče s keramiko v poznoantični cisterni. časovna opredelitev teh najdb za zdaj kaže na čas med koncem 7. in začetkom 9. stoletja. Možni sta tudi dve krajši zgodnjesrednjeveški fazi, prva nedolgo po opustitvi poznoantične naselbine ter druga v času okrog leta 800. Na prvo kažejo nekaj keramičnega gradiva ter grobova 18 in 21, na drugo pa nekaj kovinskih predmetov (Tonovcov grad. Najdbe, t. 8: 22; 26: 1, 6-8). Podobne poznamo tudi z drugih slovenskih poznoantičnih višinskih naselbin, kjer predstavljajo še ne dobro pojasnjen zgodnjekarolinški horizont (Knific 2007). 2.2.5 SREDNJI VEK Tudi iz časa po 9. stoletju je bilo na hribu najdenih nekaj predmetov (Tonovcov grad. Najdbe, t. 4: 25), ki morda nakazujejo poselitev Tonovcovega gradu tudi v srednjem veku. V ta čas morda sodijo tudi še neraziskani ostanki arhitekture na zahodnem robu naselbine (objekt št. 13), ki bi lahko predstavljali ruševine srednjeveškega stolpa. 1999, 291-292). At Tonovcov grad the remains from this period were found under buildings 1 and 2. The phase is also well represented by pottery, metal and coin finds discovered in the later layers. No Late Antiquity 1 layers were found in the area of the ecclesiastical complex and the water reservoir (cistern). LATE ANTIQUITY PHASE 2 (LA 2) The Late Antiquity phase 2 (LA 2) is even richer and better preserved than the first one. It is assumed that it spanned from the end of the 5th to the beginning of the 7th century. Most of the researched buildings in the settlement (buildings 1 and 2, the ecclesiastical complex, water reservoir) can be dated into this period, as can most of the metal and ceramic small finds. 2.2.4 EARLY MEDIEVAL PERIOD (EMP) So far the Early Medieval settlement period is not very well understood. Some small finds can be dated into it, as can some of the layers in building 1, the fireplace with some pottery in the main church, grave 21 in the area between the main and south church, the group of graves or at least one of them (grave 18) behind the south wall of building 2 and the fireplace with pottery within the Late Antique water reservoir. It seems that these finds originated from between the end of the 7th and beginning of the 9th century. Two shorter Early Medieval phases are also possible, the first shortly after the Late Antique settlement was abandoned and the second sometime around the year 800. The first is indicated by a few pieces of pottery and graves 18 and 21, while the second is indicated by several metal objects (Tonovcov grad. Finds, Pls. 8: 22; 26: 1, 6-8). Similar finds were discovered at other Late Antique hilltop settlements in Slovenia, where they formed the as yet poorly understood Early Carolingian horizon (Knific 2007). 2.2.5 MEDIEVAL PERIOD Certain objects (Tonovcov grad. Finds, Pl. 4: 25) found on the hill can be dated after the 9th century and they might indicate that some form of settlement existed at Tonovcov grad also in the Medieval period. The uninvestigated architectural remains on the west of the settlement (building 13), which could represent the ruins of a Late Medieval tower, could belong into this period. 2.3 stavba 1 z okolico 2.3 building 1 and its surroundings Za izkopavanje leta 1994 je bilo izbrano območje zahodno od vhoda v naselje, na razmeroma ravnem, le rahlo uleknjenem delu pod strmo skalno steno, ki se dviguje proti osrednjemu platoju Tonovcovega gradu. Tu sta bila že pred začetkom izkopavanj na površini vidna ruševina in delno tudi tloris velike zidane stavbe (stavba 1). V njen jugovzhodni vogal je bil v preteklosti (glej pogl. 1.1.4) narejen vkop, ki je razkril tudi del južnega zidu. Geološko osnovo v večini izkopnega polja predstavlja skala, katere površina je močno razgibana. Skala je na nekaterih mestih zelo preperela, vmesne prostore pa zapolnjuje plast trde oranžne gline. Teren pada od juga proti severu in od zahoda proti vzhodu. Izkopavanja, ki so se nadaljevala tudi v letu 1995, so zajela 480 m2 velik prostor (sl. 2.2). Poleg ostankov zidane stanovanjske stavbe (stavba 1) je bil raziskan tudi obsežen prostor okrog nje. Izkopno polje je segalo do ostankov sosednjih objektov (stavbe 10, 22, 26; sl. 1.7), zajelo pa je tudi strmo pobočje brez zidanih objektov nad stavbo 1, vse do skalne stene, ki se dviguje do platoja s cerkvami. Pod ostanki stavbe 1 je bilo dokumentiranih več faz poselitve iz časa pred njeno izgradnjo, v ruševinskih plasteh pa aktivnosti iz časa, ko je bila stavba že delno porušena. Najstarejša poselitev tega območja sodi v prazgodovino. zanesljivi naselbinski sledovi iz rimskega obdobja na območju izkopnega polja stavbe 1 niso During the 1994 excavations the area west of the settlement entrance was explored. This area lies on a relatively flat, only faintly curved part under the steep rock face that rises towards the central plateau of tonovcov grad. Even before the excavations commenced the ruin and the contour of the ground plan of a large stone building (building 1) could be seen. In the past the southeast corner was damaged and this intervention revealed a part of the south wall (see chapter 1.1.4). Throughout most of the excavation area the geological base is represented by bedrock, the surface of which is extremely diverse. In certain spots the rock has disintegrated strongly, and the spaces that appeared in between became filled by layers of compact orange clay. The terrain drops from the south to the north and from the west to the east. The excavations that continued in 1995 covered an area measuring 480 m2 in size (Fig. 2.2). Alongside the remains of the stone building that was used as the living quarters (building 1) a large area surrounding it was also researched. The excavation area reached all the way to the remains of the neighbouring structures (buildings 10, 22 and 26; Fig. 1.7) and included the steep slope (without any stone structures) above building 1 - all the way to the rock face that reaches the plateau with the churches. Under the remains of building 1 a number of settlement phases from the time before the building was Tab. 2.1: Stavba 1. Preglednica stratigrafskih enot (SE). Tab. 2.1: Building 1. Table of startigraphic units (SU). Opredelitev / Definition SE / SU Sterilna / Sterile 40 Prazgodovina / Prehistory 33, 39 PA 1 / LA 1 04, 16, 17, 20, 29a, 30, 31, 32, 36=68, 36a, 36b, 51, 53, 54, 57, 66, 68, 75, 76, 77, 78; zidovi /walls 13, 14, 15 PA 1/ PA 2 / LA 1/ LA 2 18, 21, 22, 24, 63, 74 PA 2 / LA 2 03, 05, 12, 14, 23=26, 25, 28, 29, 35a, 50, 55, 56, 62, 64, 67; zidovi / walls 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 Rušenje / Destruction 01, 08, 11, 13 ZSV / EMP 09, 10 Premešano / Mixed 02, 06, 34 Sl. 2.2: tloris stavbe 1 z mejo izkopnega polja in mrežo kvadrantov. M. = 1:200. Fig. 2.2: Ground plan of building 1 with the excavation area and quadrant grid. Scale = 1:200. bili ohranjeni (glej pogl. 2.2.2). Ponovno je območje intenzivno poseljeno v pozni antiki, ko lahko ločimo dve poselitveni fazi, to je prvo in drugo poznoantično fazo (PA 1 in PA 2). Najbolje ohranjena in najbogatejša je druga poznoantična faza, v katero sodijo tudi ostanki zidane stavbe. v stavbnih ruševinah so bili ohranjeni tudi sledovi zgodnjesrednjeveškega obdobja (tab. 2.1). erected were documented, and the destruction layers show activities from the time when the building was already partially demolished. The earliest settlement on this area can be dated to the prehistoric times. No reliable settlement traces from the Roman period were found within the excavated area (see chapter 2.2.2). The area was settled again the Late 2.3.1 PRAZGODOVINA Nad geološko osnovo je bila na nekaterih mestih Izkopnega polja najdena plast s prazgodovinskimi ostanki. Najizrazitejša je bila na severni strani izkopnega polja, v kvadrantih 617, 618, 619, 668 in 669, kjer je bila skalna osnova najnižja in so bile kulturne plasti najdebelejše. Skalno osnovo je prekrivala plast trde oranžne, kulturno sterilne gline (SE 40). Nad njo je ležala tanka (10-20 cm) plast trde oranžno-rjave gline (SE 33), ki je vsebovala živalske kosti, drobce oglja in kamnite artefakte (sl. 2.3). Na njeni zgornji površini, na meji s SE 32, so se že pojavljale antične najdbe. Kamniti artefakti so bili najdeni tudi še v SE 32, ki pa je bila glede na ostale najdbe že antična. ZID 4/ WALL 4 396,50 m □ 01 rjava mehka prst z velikimi kamni / brown loose soil with large stones 08 plast odpadlega ometa / layer of fallen off plaster 23 črna prst s kamni / black soil with stones 24 sivo-rjava prst z gruščem / greyish-brown soil with gravel 28 sivo-rjav pesek / greyish-brown sand 29a temno rjava prst s kamni / dark brown soil with stones 32 rjava peščena prst / brown sandy soil 33 oranžno-rjava glina / orange-brown clay 40 oranžna glina / orange clay Sl. 2.3: Stavba 1, presek 4 pravokotno na zid 4. M. = 1:50. Fig. 2.3: Building 1, section 4 at a right angle to wall 4. Scale = 1:50. V osrednjem delu izkopnega polja (kv. 717, 718, 719, 767, 768) je bilo nekaj prazgodovinskih ostankov ohranjenih le v skalnih razpokah, zapolnjenih s trdo oranžno glino (SE 39). V južnem in zahodnem delu izkopnega polja, kjer se skalna osnova močno dvigne, prazgodovinski ostanki niso bili najdeni. 2.3.2 PRVA POZNOANTIČNA FAZA (PA 1) Nad prazgodovinsko plastjo oziroma nad geološko osnovo so ležale kulturne plasti, ki predstavljajo ostanek prve poznoantične faze, to je faze pred gradnjo Antiquity. This period could be divided in two settlement phases: late Antiquity phase 1 (lA 1) and late Antiquity phase 2 (lA 2) . The best preserved and richest in finds is late Antiquity phase 2, which also includes the remains of the stone building. Traces of the Early Medieval period were preserved within the destruction layers (Tab. 2.1). 2.3.1 PREHISTORY In some spots a layer with prehistoric remains was found above the geological base. This was most pronounced on the north side of the excavation area in quadrants 617, 618, 619, 668 and 669, where the bedrock was the lowest and the cultural layers the thickest. The bedrock was covered by a layer of compact orange clay that did not show any cultural traces (SU 40). This was covered by a thin (10-20 cm) layer of compact, orange-brown clay (SU 33) that contained animal bones, charcoal particles and stone artefacts (Fig. 2.3). On its upper surface, on the border with SU 32, finds dating to the Roman period were discovered. Stone artefacts were also found in SU 32, however the remaining finds date this layer into the Roman period. In the central part of the excavation area (qus. 717, 718, 719, 767, 768) the few prehistoric remains were preserved merely in the rock crevices that were filled with compact orange clay (SU 39). No prehistoric remains were discovered in the southern or western part of the excavation area where the bedrock rises sharply. 2.3.2 LATE ANTIQUITY PHASE 1 (lA 1) On top of the prehistoric layer or on top of the geological base lay cultural layers that represent the remnants of late Antiquity phase 1, i.e. the phase that pre-existed building 1. During the construction of building 1 (at the beginning of late Antiquity phase 2) the layers were severely damaged, however we can still ascertain that these remains - similar to the prehistoric ones - were also based in the northern and eastern part of the excavation area. Remains were documented in the interior of the later building as well as on the outside of it (and under its walls), however they were severely damaged during construction. In the eastern part of the excavation area a number of cultural layers (SU 30, 31 and 32) that included finds from the Antiquity were discovered above the prehistoric layer SU 33. SU 32 ran under wall 4 also in the interior of the later object (Figs. 2.3,2.4). SU 29a - a layer of loose soil with a strong presence of cultural finds and larger stones (Fig. 2.3) - was located on top of SU 32. This layer also ran under wall 4 (Figs. 2.3, 2.4), however it was not preserved in the interior of the later building. ZID 4 / WALL 4 397,00 m 29a temno rjava prst s kamni / dark brown soil with stones 32 rjava peščena prst / brown sandy soil 40 oranžna glina / orange clay ■ omet/plaster m skala / bedrock Sl. 2.4: Stavba 1, zid 4 - zunanje lice. M. = 1:50. Fig. 2.4: Building 1, wall 4 - exterior. Scale = 1:50. stavbe 1. Plasti so bile ob gradnji stavbe 1 na začetku druge poznoantične faze zelo uničene, kljub temu pa lahko ugotovimo, da so bili tudi ti ostanki - podobno kot prazgodovinski - strnjeni v severnem in vzhodnem delu izkopnega polja. Dokumentirani so bili tako v notranjosti kot zunaj kasnejše stavbe, pa tudi pod njenimi zidovi, vendar so bili ob gradnji zelo poškodovani. v vzhodnem delu izkopnega polja je nad prazgodovinsko SE 33 ležalo več kulturnih plasti (SE 30, 31 in 32), ki so že vsebovale antične najdbe. SE 32 se je nadaljevala pod zidom 4 tudi v notranjost kasnejšega objekta (sl. 2.3,2.4). Nad SE 32 je ležala SE 29a, to je rahla, močno kulturna prst, v kateri so bili tudi večji kamni (sl. 2.3). Tudi ta plast se je nadaljevala pod zid 4 (sl. 2.4), v notranjosti kasnejše stavbe pa ni bila več ohranjena. ob zunanjem vogalu zidov 3 in 4 so bili ohranjeni ostanki dveh zelo uničenih zidov, to sta zidova 13 in 14 (sl. 2.2), od katerih je ostala samo spodnja linija v vrsto postavljenih kamnov brez sledov malte ali ometa. Zid 13 je potekal približno pravokotno na zid 4, vendar z njim ni bil neposredno povezan, saj je bila med njima približno 20 cm široka reža. Bil je dolg približno 3 m, nanj se je pravokotno navezoval zid 14, ki je bil v dolžino ohranjen samo še približno 2 m. Spodnja linija zidu 13 je bila 30-40 cm niže kot spodnja linija zidu 4. v osrednjem delu izkopnega polja (kv. 767 in delno kv. 717) je plasti prve poznoantične faze zelo poškodovala gradnja stavbe 1. Tu so nad glineno osnovo SE 40 ležale zaplate rjave gline (SE 36; sl. 2.5,2.6). Plast je bila močno kulturna, v mkv. 718/D4 sta bili vanjo vkopani dve jami nepravilne oblike (SE 36a in SE 36b). Jama 36a je bila zapolnjena s črno mehko zemljo, vsebovala pa je tudi opeko, keramiko in živalske kosti. Tik ob njej je ležala strnjena plast opeke, poleg nje pa keramika in kosti. Jama 36b je bila zapolnjena z zemljo, vendar je bila brez najdb. The remains of two severely destroyed walls (wall 13 and wall 14) were discovered on the exterior of the corner where walls 3 and 4 meet (Fig. 2.2). Only the lower line of stones was preserved, without any traces of mortar or plaster. Wall 13 ran roughly at a right angle to wall 4, however it had no direct contact with it, for there was an approximately 20 cm wide gap between the two. It was approximately 3 metres long and wall 14 (of which roughly a mere 2 metres in length was preserved) was attached to it at a right angle. The bottom row of wall 13 was 30-40 cm lower than the lower line of wall 4. The Late Antiquity 1 layers in the central part of the excavation area (qu. 767 and partially qu. 717) were severely damaged during the construction of building 1. Patches of brown clay (SU 36) were discovered above the clay base (SU 40; Figs. 2.5, 2.6). This layer had a strong cultural presence, and mqu. 718/D4 revealed two irregular shaped pits (SU 36a and SU 36b). Pit 36a was filled with loose black soil that included bricks, pottery and animal bones. A composite layer of bricks was discovered next to it and alongside this lay pottery and animal bones. Pit 36b was filled with soil but was void of any finds. SU 36 ran under wall 1 of the later building (Figs. 2.5,2.6) also on the north of the excavation area, in quadrants 716 and 666 (Fig. 2.7). Under wall 1 another two cultural pits (SU 77 and SU 78) could be observed, both of which were dug all the way through the cultural layers right to the bedrock (Fig. 2.5). Wall 15 stood in the northern part of the excavation area, on SU 36 (Figs. 2.2, 2.7). The wall was 5.4 metres long and 0.5 metres wide, and merely a single row was preserved in height. Its width was covered by one large or two slightly smaller stones, without an intermediate gravel layer. Modest mortar remains were discovered amongst the stones. ZID 2/ WALL 2 397,00 m ZID 6/WALL 6 24 sivo-rjava prst z gruščem / greyish-brown soil with gravel 36 rjava glinena prst / brown clayey soil 40 oranžna glina / orange clay 77 črna mehka prst / black loose soil 78 rjava mehka prst / brown loose soil ■ omet / plaster keramika / pottery Q oglje / charcoal HI skala / bedrock Sl. 2.5: Stavba 1, zid 1 - notranje lice in vhod. M. = 1:50 Fig. 2.5: Building 1, wall 1 - interior and threshold. Scale = 1:50. ZID 1 / WALL 1 397,00 m ZID 7 / WALL 7 iTiV sgSfr 397,00 m 123 24 sivo-rjava prst z gruščem / greyish-brown soil with gravel ■ 36 rjava glinena prst / brown clayey soil ■ 40 oranžna glina / orange clay ■ omet / plaster la skala / bedrock Sl. 2.6: Stavba 1, zid 1 - zunanje lice. M. = 1:50 Fig. 2.6: Building 1, wall 1 - exterior. Scale = 1:50. SE 36 je potekala pod zidom 1 kasnejše stavbe (sl. 2.5,2.6) tudi na severni del izkopnega polja v kvadrante 716 in 666 (sl. 2.7). Pod zidom 1 sta bili vidni še dve kulturni jami, SE 77 in SE 78, ki sta bili obe vkopani skozi kulturne plasti do skalne osnove (sl. 2.5). v severnem delu izkopnega polja je na SE 36 ležal zid 15 (sl. 2.2, 2.7). Bil je 5,4 m dolg in 0,5 m širok, v višino je bil ohranjen samo še v eno vrsto. V širino je bil zgrajen iz dveh ali celo samo iz enega večjega kamna, brez vmesne plasti drobirja. Med kamni so bili ohranjeni skromni ostanki malte. Alongside walls 1, 4 and 5 - on the outer side of the later building 1 (above SU 36) - lay a thick cultural layer (SU 24), into which walls 1 and 4 as well as a part of wall 5 were dug. The SU 24 remains were also visible at some other locations under the building walls (wall 1: Figs. 2.5, 2.6; wall 7: Fig. 2.8), while in the interior the entire layer was removed when the construction started. In some places alongside the wall SU 24 reached as high as 0.5 metres. It also covered the remains of wall 15. The layer declined from the west to the east, in the same H 36 rjava glinena prst / brown clayey soil £23 51 rumen pesek / yeliow sand □ 56 siva mehka prst / grey loose soii H 57 oranžno-rjava glina / orange-brown clay m skaia / bedrock Sl. 2.7: Stavba 1 - zunaj, kv. 666, 667 in 716, planum 3. M. = 1:50 Fig. 2.7: Building 1- exterior, qus. 666, 667 and 716, planum 3. Scale = 1:50. ob zidovih 1, 4 in 5 na zunanji strani kasnejše stavbe je nad SE 36 ležala debela kulturna plast SE 24, v katero sta bila vkopana tudi zidova 1 in 4 ter del zidu 5. ostanki SE 24 so bili vidni tudi na nekaterih mestih pod zidovi stavbe (zid 1: sl. 2.5, 2.6; zid 7: sl. 2.8), medtem ko je bila v notranjosti ob izkopu gradbene jame popolnoma odstranjena. SE 24 je ob zidovih kasnejše stavbe na nekaterih mestih segala tudi do 0,5 m visoko. Prekrivala je tudi ostanke zidu 15. Plast je padala od zahoda proti vzhodu, tako kot pada tudi geološka osnova. Zgornji nivo SE 24 ob severozahodnem vogalu je bil na pribl. 397,35 m n. m., ob vhodu v stavbo pa že 396,55 m n. m. Na delu, kjer se ob zidu 5 skalna osnova močno dvigne, se je SE 24 izklinila in se spet pojavila ob severovzhodnem vogalu. ob zidu 4, kjer je prekrivala SE 29a, je bil njen zgornji nivo približno na višini 396,5 m n. m. (sl. 2.3). ZID 1 / WALL 1 ZID 7 / WALL 7 397,00 m 397,00 m Prag/ Threshold EEE 06 malta/mortar EI3 24 sivo-rjava prst z gruščem / greyish-brown soii with gravei H omet / plaster H skala / bedrock Sl. 2.8: Stavba 1, zid 7 - notranje lice in stik z zidom 1. M. = 1:50 Fig. 2.8: Building 1, wall 7 - interior and contact with wall 1. Scale = 1:50. Sl. 2.9: Stavba 1, presek 1. M. = 1:50. Fig. 2.9: Building 1, section 1. Scale = 1:50. Za zidovoma 3 in 11 se skalna osnova strmo dvigne. Tu je kulturno plast prve poznoantične faze predstavljala oranžna ilovica SE 04, ki je ležala neposredno na skalni osnovi (sl. 2.9). Na zahodnem delu izkopnega polja in na območju kasnejšega prizidka stavbe 1 (zidovi 8-11) plasti, ki bi pripadale prvi poznoantični fazi, ni bilo. Tudi tu je skalna osnova precej višja kot na osrednjem delu in ruševinske plasti stavbe 1 so ležale neposredno na skali. 2.3.3 DRUGA POZNOANTIČNA FAZA (PA 2) Nad ostanki prve poznoantične faze je bil postavljen zidan bivalni objekt (stavba 1), delno vkopan vanje. Sestavljala sta ga glavni prostor pravokotnega tlorisa in nanj prizidan manjši prostor (pril. 1; sl. 2.2, 2.10). Stavba je bila usmerjena približno jugozahodno-severovzhodno. Vhod v osrednji prostor je bil s severozahodne strani. Bil je zaščiten z dvema, pravokotno na zidova 1 in 5 postavljenima zidcema (zidova 6 in 7), dolgima 1,8 m in širokima 0,6 m, ki sta služila kot vetrolov. Zidova 1 in 5 je povezoval kvalitetno izdelan prag (sl. 2.10,2.11), katerega zgornja površina je ležala na višini približno 396,6 m n. m. Na jugozahodni zid glavnega prostora (zid 2) je bil prizidan stranski prostor (prizidek), ki je bil pribl. 1,2 m krajši od glavnega prostora (pril. 1; sl. 2.12). Tudi vhod v prizidek je vodil s severozahodne strani, ostanki praga pa tu niso bili najdeni. way as the geological base. In the northwest corner the upper level of SU 24 was approximately at 397.35 m a.s.l., while at the wind shield it was already at 396.55 m a.s.l. In the part where the bedrock rises sharply alongside wall 5, Su 24 disappears only to reappear in the northeast corner. Alongside wall 4, where it covered SU 29a, its upper level was approximately at an altitude of 396.5 m a.s.l. (Fig. 2.3). The bedrock rises sharply behind walls 3 and 11. Here the cultural layer belonging to late Antiquity phase 1 was represented by orange clay (SU 04) which lay directly on the bedrock (Fig. 2.9). Not a single layer b elonging to late Antiquity phase 1 was discovered in the western part of the excavation area and in the area that was later covered by the outhouse. At this location the bedrock is also much higher than in the central part and the building 1 destruction layers were discovered directly on the rock. 2.3.3 LATE ANTIQUITY PHASE 2 (LA 2) Above the Late Antiquity 1 remains, partially dug into them, stood a stone building (building 1). It consisted of the main room with a rectangular ground plan and a smaller outhouse (Insert 1; Figs. 2.2,2.10). The orientation of the building was approximately southwest-northeast. The entrance into the main room was located on the northwest and was protected by two small walls (walls 6 and 7). The walls (1.8 m long and 0.6 m wide) were set at a right angle to walls 1 and 5 and Sl. 2.10: Stavba 1, glavni prostor in del prizidka. Pogled s severozahoda. Fig. 2.10: Building 1, main room and part of the outhouse. A view from the northwest. Sl. 2.11: Stavba 1, vhod s pragom in vetrolovom. Fig. 2.11: Building 1, entrance with threshold and wind shield. ziDovi (pril. 1) zidovi glavnega prostora stavbe 1 so bili zidani iz apnenčevih lomljencev. večinoma so bili grajeni v dveh vrstah, široki med 55 in 60 cm. Prostor med vrstami je bil zapolnjen z gruščem in malto. Na mestih, kjer zidovi niso bili postavljeni na skalno osnovo, ampak na mehkejše plasti, so bili temelji grajeni iz večjih kamnov, postavljenih v eni vrsti. Lica so bila zamazana z malto, tako da so bili vidni samo vrhovi kamnov. served as a wind shield. Walls 1 and 5 were connected by a high quality threshold (Figs. 2.10,2.11), the upper surface of which stood at approximately 396.6 m a.s.l. The adjacent room (the outhouse) was added to wall 2 of the main space, and it was approximately 1.2 m shorter than the main room (Insert 1; Fig. 2.12). The entrance into the outhouse was also located on the northwest, however no threshold remains were found. WALLS (Insert 1) The walls of the main room of building 1 were constructed from limestone quarry stones. They were mostly constructed from two rows of stones, and were between 55 and 60 cm thick. The space between the rows was filled with gravel and mortar. Wherever the walls were not placed onto the bedrock, but onto looser layers, the foundations were constructed from larger stones, set in a single row. The side wall surfaces were covered in mortar so that only the top of the stones could be seen. The walls of the outhouse were built almost entirely without mortar and are narrower (approx. 50 cm), constructed from two or sometimes even a single row of stones, without a layer of gravel in between. Sl. 2.12: Stavba 1, prizidek. Pogled s severovzhoda. Fig. 2.12: building 1, outhouse. A view from the northeast. Zidovi stranskega prostora so bili zidani skoraj brez malte in so ožji (okrog 50 cm), grajeni v dveh ali pa tudi le v eni vrsti, brez vmesne plasti grušča. Zid 1 (sl. 2.5,2.6,2.13) Zid 1 je dolg 4,4 m in s pragom povezan z zidom 5. Na notranji strani poteka od vrha do temeljev poševna razpoka. Ob stiku z zidom 2 je bil postavljen na skalno osnovo, ob vhodu pa na plast SE 24, ki predstavlja ostanek starejše poselitve, v katero sta bili vkopani še dve močno kulturni jami SE 77 in SE 78, vidni samo na notranji strani stavbe (sl. 2.5). Ob vhodu zid 1 temelji na vrsti večjih, nekoliko izstopajočih kamnitih lomljen-cev (sl. 2.13), na njegovo zunanje lice pa je pravokotno prizidan zid 7 (vetrolov). Sl. 2.13: Stavba 1, zid 1 - notranje lice. Fig. 2.13: Building 1, wall 1 - interior. Zid 2 (sl. 2.14, 2.15) Zid skoraj v celi dolžini stoji na plasteh prve poznoantične faze SE 36 in SE 20. temelje sestavlja vrsta večjih neometanih kamnitih lomljencev, katerih zgornja površina je zravnana s plastjo malte, nanje pa so postavljeni manjši kamni zidu. V južnem delu je v dolžini 1,4 m med temeljem in samim zidom s ploščatimi kamni izoblikovana približno 10 cm široka reža (sl. 2.15). Wall 1 (Figs. 2.5,2.6,2.13) Wall 1 is 4.4 m long and is connected to wall 5 with a threshold. On the inner side a diagonal crack runs from the top to the very foundations. At the contact with wall 2 it was placed directly onto the bedrock, while at the entrance it stood on SU 24 which represents the remains of an older settlement into which two pits (SU 77 and SU 78) were dug in, both with a strong presence of cultural remains and both of which can be seen only in the building interior (Fig. 2.5). At the entrance wall 1 is based on a series of large, slightly protruding quarry stones (Fig. 2.13), while wall 7 (the wind shield) is attached at a right angle to the outer side. Wall 2 (Figs. 2.14, 2.15) This wall stands on Late Antiquity 1 layers SU 36 and SU 20 almost along its entire length. The base is represented by a line of large non-plastered quarry stones, the upper surface of which is levelled by a layer of mortar, onto which smaller stones were placed. In the south a series of flat stones formed a 1.4 metre long and roughly 10 cm wide fissure between the base and the wall (Fig. 2.15). The stones above the base were covered by a thick layer of plaster, thus only the tops of the stones could be seen in some places. At the point where the outhouse was attached to it the outer side of wall 1 was also covered in plaster. Wall 3 (Figs. 2.16-2.18) This is the longest (10.8 m) single wall of building 1. It was built on a very uneven terrain, for in the central part it stands on a rock that rises approximately 1.2 m WALL 3 p^^^ 397,00 m 1 / , a pop WALL 1 397,00 m 1 m 20 oranžno-rjava glina / orange-brown clay 36 rjava glinena prst / brown clayey soil 40 oranžna glina / orange clay omet / plaster skala / bedrock Sl. 2.14: Stavba 1, zid 2 - notranje lice. M. = 1:50. Fig. 2.14: Building 1, wall 2 - interior. Scale = 1:50. --CM CM ^ n| Sl. 2.15: Stavba 1, stik zidov 2 in 3. Odprtina med kamni v zidu 2 in iz malte oblikovana polička v kotu. Fig. 2.15: Building 1, contact between walls 2 and 3. The fissure between the stones in wall 2 and the mortar shelf in the corner. Kamni zidu nad temelji so bili ometani z veliko količino malte, tako da so na nekaterih mestih vidni le vrhovi kamnov. Ometana je bila tudi zunanja stran zidu 1, kjer je bil nanj prizidan stranski prostor. Zid 3 (sl. 2.16-2.18) Je najdaljši (10,8 m) sklenjeni zid stavbe. Zgrajen je bil na zelo neravnem terenu, saj na osrednjem delu stoji na skali, ki se pribl. 1,2 m visoko dviga nad siceršnjo osnovo terena, takšna pa je tudi razlika med spodnjim nivojem zidu na osrednjem in severovzhodnem delu. Zid je bil zgrajen dokaj neenakomerno. Ob stiku z zidom 4 je v dolžini približno 3,5 m temeljen na eni vrsti večjih kamnov, ki stojijo na sterilni glini (SE 40). Ob stiku z zidom 2 stoji na kulturni plasti prve poznoantične faze SE 04. Tu je spodnja vrsta kamnov porušena navzdol. Na osrednjem delu, kjer sega nivo skale najvišje, posebnega temeljenja ni, zid je na tem mestu zgrajen iz dokaj pravilno izoblikovanih kamnov v poravnanih vrstah (sl. 2.16, 2.17). Na severovzhodnem delu, kjer nivo skale močno pade, je v dolžini približno 2,4 m v treh vrstah opazno močno temeljenje (sl. 2.16). Blizu vogala z zidom 4 je nad nivojem temeljnih kamnov izoblikovana polička iz malte (sl. 2.18). Prostori med kamni zidu na notranji strani so bili ometani precej manj kot pri zidu 2, vendar ni povsem jasno, ali je to prvotno stanje ali stanje ohranjenosti ometa. Zunanja stran zidu 3, ki gleda proti strmemu pobočju, ni bila ometana. i I A rfMM J^fVfVUj^ rVnl g ow Sl. 2.16: Stavba 1, zid 3 - notranje lice. M. = 1:50. Fig. 2.16: Building 1, wall 3 - interior. Scale = 1:50. Sl. 2.17: Stavba 1, zid 3 - notranje lice. Detajl osrednjega dela zidu. Fig. 2.17: Building 1, wall 3 - interior. Detail of the central part. Sl. 2.19: Stavba 1, zid 4 - zunanje lice. Fig. 2.19: Building 1, wall 4 - exterior. Sl. 2.18: Stavba 1, stik zidov 3 in 4. Iz malte oblikovana polička v kotu. Fig. 2.18: Building 1, contact between walls 3 and 4. The mortar shelf in the corner. Sl. 2.20: Stavba 1, zid 5 - zunanje lice. Fig. 2.20: Building 1, wall 5 - exterior. Zid 4 (sl. 2.4, 2.19) Zid je bil vkopan v mehko kulturno plast SE 29a skoraj do pod njo ležeče rjave peščene plasti SE 32 (sl. 2.3). Temelji zidu so bili grajeni iz velikih kamnitih lomljencev, kar je še posebej izrazito ob stiku z zidom 5 (sl. 2.19). Tudi pri zidu 4 je na notranji strani (podobno kot pri zidu 2) v dolžini približno 1,5 m med temelji in samim zidom opazna vodoravna reža, ki pa je tukaj manj izrazita. Prostor med kamni na zunanji in notranji strani zidu je bil obilno ometan. Zid 5 (sl. 2.20) Zid je bil skoraj v celoti postavljen na skalo, razen ob vhodu v stavbo, kjer je v dolžini pribl. 1 m temeljen na glineno osnovo. Nivo skale ob stiku z zidom 4 močno pade in na tem mestu je bil zgrajen močan temelj iz velikih kamnitih blokov (glej tudi opis zidu 4). Ob vhodu v stavbo je bil pravokotno na zid 5 prizidan zid vetrolova (zid 6). above the clay base, which also represents the distance between the lower level of the wall in the central and northeastern part. The wall was built rather unevenly. At the contact with wall 4, approximately 3.5 m in length, the foundations are represented by a single line of larger stones that are positioned on sterile clay (SU 40). At the contact with wall 2 it stands on the cultural layer SU 04 that belongs to late Antiquity phase 1. At this point the lower line of stones has crumbled downwards. In the central part of the wall, where the rock level is at its highest, there are no special foundations, and the wall is built from relatively well formed stones that run in a straight line (Figs. 2.16,2.17). In the northeast, where the base drops considerably, strong foundations measuring approximately 2.4 metres in length and made of three rows of stones were discovered (Fig. 2.16). A small mortar shelf was created close to the corner with wall 4, above the level of the base stones (Fig. 2.18). On the inner side of wall 3 the spaces between the stones were plastered by a much thinner layer of plaster Sl. 2.21: Stavba 1, detajl vhodnega dela, zid 6 - notranje lice in stik z zidom 5 in pragom. Fig. 2.21: Building 1, detail of the entrance, wall 6 - interior and contact with wall 5 and threshold. Sl. 2.22: Stavba 1, zid 7 - notranje lice in stik z zidom 1 in pragom. Fig. 2.22: Building 1, wall 7 - interior and contact with wall 1 and threshold. Zid 6 (sl. 2.2, 2.11, 2.21) Severovzhodni zid vetrolova je bil prizidan pravokotno na zid 5, tako da je bil zamaknjen pribl. 20 cm proti notranjosti vhoda (pril. 1; sl. 2.2, 2.11, 2.21). Temeljen je na skali oziroma skalni preperini, in sicer pribl. 20 cm globlje kot zid same stavbe. V višino so bile ohranjene le dve do tri vrste. Zid 7 (sl. 2.8, 2.11, 2.22, 2.29) Prizidan je bil pravokotno na zid 1 tako, da je bil prib. 20 cm zamaknjen proti notranjosti vhoda. Tudi zid 7 je bil temeljen pribl. 20 cm nižje kot sam zid stavbe in prag, postavljen pa je bil na SE 24, to je kulturno plast prve poznoantične faze (sl. 2.8). Temelj je bil zgrajen iz večjih, sam zid pa iz manjših, dokaj pravilno oblikovanih ploščatih kamnov. Stiki med njimi so bili obilno prekriti z ometom. than at wall 2, however it is not absolutely clear whether this was the original state or is this merely the state of preservation. The outer side of wall 3 (looking towards the steep slope) was not covered in plaster. Wall 4 (Figs. 2.4, 2.19) The wall was dug into the loose cultural layer (SU 29a) almost all the way to the brown sand layer (SU 32) that lies underneath (Fig. 2.3). The wall foundations were constructed from large quarry stones, which is especially noticeable at the contact with wall 5 (Fig. 2.19). On the inner side of wall 4 (similar as is the case with wall 2) a horizontal fissure was discovered between the foundation and the wall itself. The fissure measures approximately 1.5 m in length, however it is not as noticeable as the fissure in wall 2. The space between the stones on the outer as well as on the inner side of the wall was covered by a thick layer of plaster. Wall 5 (Fig. 2.20) Almost the entire wall was positioned on the bedrock, except for the entrance into the building that was positioned on a clay base (approximately 1 metre in length). At the contact with wall 4 the bedrock declines sharply and thus a strong foundation from large stone blocks was constructed (see the description of wall 4). At the entrance into the building the wall of the wind shield (wall 6) was built at a right angle to wall 5. Wall 6 (Figs. 2.2, 2.11, 2.21) The wall belonging to the wind shield was built at a right angle to wall 5 in such a way that it covered approximately 20 cm of the inner side of the entrance (Insert 1; Figs. 2.2, 2.11, 2.21). It was based on the rock approximately 20 cm below the wall of the building itself. only two to three rows were preserved in height. Wall 7 (Figs. 2.8, 2.11, 2.22, 2.29) This wall was built at a right angle to wall 1, approximately 20 cm towards the entrance interior. The foundations of wall 7 were also approximately 20 cm lower than the wall of the building and the threshold, and stood on the cultural layer SU 24, which belonged to the first Late Antiquity phase (Fig. 2.8). The foundation was constructed from larger, and the wall from smaller, relatively well worked flat stones. The contacts between them were covered by a thick layer of plaster. Sl. 2.23: Stavba 1, zid 8 in stik z zidom 2. Fig. 2.23: Building 1, wall 8 and contact with wall 2. Sl. 2.24: Stavba 1, del zidu 10 - notranje lice. Fig. 2.24: Building 1, part of wall 10 - interior. Zid 8 (sl. 2.23) Zid je bil prizidan na zid 2 glavnega prostora pribl. 1 m od vogala zidov 1 in 2 (sl. 2.2). Bil je zgrajen potem, ko je bil zid glavnega prostora že ometan. 1 m po stiku z zidom 2 je prekinjen z vhodom. Brez posebnega temeljenja je bil postavljen na SE 14 približno 30 cm višje kot zid 2 glavnega prostora. Zid 8 ni bil ometan, tudi v njegovi notranjosti ni bilo opaziti nobenih sledov malte. Wall 8 (Fig. 2.23) The wall was attached to wall 2 of the main room roughly 1 metre away from the corner between walls 1 and 2 (Fig. 2.2). It was constructed once wall 2 was already covered by a layer of plaster. 1 metre after the connection with wall 2 the wall was cut off by the entrance. The wall was built on SU 14 approximately 30 cm higher than wall 2 and did not have any special foundations. Wall 8 was not covered in plaster, and no traces of mortar were found on the inner side. Zid 9 Zid je dolg 1,4 m in podobno kot zid 8 neometan, brez sledov malte in posebnega temeljenja. V kotu, ki ga tvori z zidom 10, je bilo ohranjeno ognjišče (ognjišče 1). Zid 10 (sl. 2.24) Zid v večini leži na skalni osnovi, ki močno pada od jugovzhoda proti severozahodu, le ob stiku z zidom 9, kjer se teren nekoliko zravna, leži na kulturni plasti SE 12 (sl. 2.24). Tudi pri zidu 10 ni bilo opaziti nobenega posebnega temeljenja, zid tudi ni bil ometan. Med kamni zidu je bilo ohranjenih nekaj skromnih ostankov malte. Na zunanji strani tik ob zidu je ležal skelet, položen neposredno na skalno osnovo in prekrit z ruševino zidu (grob 1). Zid 11 (sl. 2.25) Zid je prislonjen na zid 2 na vogalu z zidom 3 (sl. 2.25). Temeljen je pribl. 40 cm plitveje kot zid glavne stavbe. Tudi na tem mestu skalna osnova močno pada, in sicer od jugozahoda proti severovzhodu. Zid je tako v Wall 9 The wall is 1.4 m long and similar to wall 8 it is not covered in plaster, it shows no traces of mortar and has no explicit foundations. A fireplace was preserved in the corner with wall 10 (fireplace 1). Wall 10 (Fig. 2.24) Most of the wall lies on the bedrock that sharply declines from the southeast to the northwest, for only at the point where it touches upon wall 9, i.e. where the terrain levels out slightly, is it positioned on the cultural layer SU 12 (Fig. 2.24). No explicit foundations or traces of plaster were noticed. A few traces of mortar were found amongst the stones in the wall. A skeleton was discovered on the outer side, right next to the wall. The skeleton was placed directly onto the bedrock and was covered by the wall ruins (grave 1). Wall 11 (Fig. 2.25) The wall reached wall 2 at the corner with wall 3 (Fig. 2.25). The foundations were approximately 40 cm Sl. 2.25: Stavba 1, del zidu 11 in stik z zidom 2. Fig. 2.25: Building 1, part of wall 11 and contact with wall 2. Sl. 2.26: Stavba 1, postament (zid 12). Pogled z jugovzhoda. Fig. 2.26: Building 1, pedestal (wall 12). A view from the southeast. večini postavljen na skalo, razen ob stiku z zidom 2, kjer je bil močan skalni padec zravnan s peščenim nasutjem. Samo na tem mestu je opaziti nekaj večjih kamnov, ki so služili za temelj. Zid ni bil ometan. Zid 12 - postament (sl. 2.26, 2.27) Na zunanji strani zidu 1 je bil pribl. 1 m oddaljen od njega postavljen zidan podstavek dimenzij 70 x 70 cm, vkopan v SE 24 (sl. 2.27). Njegovi temelji segajo v globino približno toliko kot temelji zidu 1, v višino so ohranjene štiri vrste kamnov, ki jih povezuje velika količina malte. Z malto je bil močno premazan tudi na vrhu. shallower than those of the wall in the main building. At this point the bedrock also declines sharply in the northeast direction. The wall is thus mainly positioned directly on the bedrock, except where it reaches wall 2, where the great decline of the bedrock was levelled out with gravel. This is the only location where a few larger stones that used to be a part of the foundations were noticed. The wall was not covered in plaster. Wall 12 - pedestal (Figs. 2.26, 2.27) On the outer side of wall 1, approximately 1 m from the wall, a 70 x 70 cm stone pedestal was dug into SU 24 x = 64 y = 70 IS 24 sivo-rjava prst z gruščem / greyish-brown soil with gravel El 35 malta / mortar ES 35a malta, mešana z zemljo in kamni / mortar mixed with soil and stones ® x = 64 y = 66 Sl. 2.27: Stavba 1 - zunaj, kv. 716, 766, planum 2. M. = 1:50. Fig. 2.27: Building 1- exterior, qus. 716, 766, planum 2. Scale = 1:50. PRAG (pril. 1; sl. 2.5, 2.11, 2.21, 2.22) Prag je povezoval zidova 1 in 5. Zgrajen je bil iz 10-15 cm debele plasti kamenja in kakovostne malte, s katero so bili kamni na delu, obrnjenem proti notranjosti hiše, tudi premazani. Po vsej dolžini je bil v malti izdelan približno 50 cm širok in 2 cm globok utor. Prag je v večjem delu temeljil na skalni osnovi. Na strani, obrnjeni proti notranjosti stavbe, je podenj ob stiku z zidom 1 segal vkop prve poznoantične faze (SE 78), ob stiku z zidom 5 pa skromni ostanki SE 24 (sl. 2.5). OGNJIŠČE 1 (sl. 2.28, 2.29) v jugovzhodnem delu glavnega prostora je bilo približno 1,2 m oddaljeno od zidu 3 postavljeno zidano ognjišče, veliko približno 1,2 x 1 m. Zgrajeno je bilo iz večjih, vodoravno položenih ploščatih kamnov, ki so bili ob robu postavljeni navpično, tako da so tvorili nizko ograjo. Pri gradnji so uporabili tudi kose opeke in del zgornjega žrmljnega kamna. Na nekaterih mestih je bil še ohranjen glinen premaz, ki je nekoč verjetno pokrival celo površino. ognjišče je bilo delno postavljeno na skalno osnovo, delno pa na glineno plast prve pozno-antične faze SE 18. Sl. 2.28: Stavba 1, ognjišče 1, pogled z zahoda. Fig. 2.28: Building 1, fireplace 1, a view from the west. (Fig. 2.27). The foundations are roughly the same depth as the foundations of wall 1, and four stone rows joined by large quantities of mortar remain in height. The top of the wall was also covered with large quantities of mortar. THRESHOLD (Insert 1; Figs. 2.5, 2.11, 2.21, 2.22) The threshold joined walls 1 and 5. It was constructed from a 10 to 15 cm thick layer of stones and high quality mortar, with which the stones that were turned towards the house interior were also covered. Along the entire length ran a roughly 50 cm wide and 2 cm deep groove into the mortar. To a greater extent the threshold was based on the bedrock. On the side turned towards the interior of the building, a pit (SU 78) belonging to Late Antiquity phase 1 reached under the threshold at the point where it touched wall 1. At the contact with wall 5 modest remains of SU 24 were found (Fig. 2.5). FIREPLACE 1 (Figs. 2.28, 2.29) A stone fireplace, approximately 1.2 x 1 m in size, was positioned in the southeast part of the main room, approximately 1.2 m from wall 3. The fireplace was constructed from large, horizontally placed flat stones, and on the edge the stones were placed vertically so that they formed a low fence. Brick parts and a part of the upper stone of a rotary-quern were also used in the construction. In some places a clay coating was preserved. It is highly likely that it covered the entire surface. The fireplace was partially positioned on the bedrock and partially on the clay layer SU 18 that belonged to the first Late Antiquity phase. FIREPLACE 2 (Insert 1) Next to the entrance in the outhouse, in the corner created by walls 9 and 10, stood a 10 cm high fireplace created from flat, heavily cracked stones and brick particles. It was constructed in a much poorer fashion than fireplace 1 in the main room. m 02 recentni vl1 Ajw 80 ® /II ^ IL'-- ■ 78 rjava prst s kamni / brown soil with stones 80 rjav grušč / brown gravel I'l;' .. .. lllllisi skala / bedrock x = 92 y = 22 Sl. 2.85: Prostor med osrednjo in južno cerkvijo. Jugovzhodni sektor, planum 7. M. = 1:50. Fig. 2.85: Space between the main and south church, southeastern sector, planum 7. Scale = 1:50. :D 1 m ■ 78 rjava prst s kamni / brown soil witii stones ■ 104 trda maltna plast 3 / hard iayer of mortar 3 ■ 106 ogije/charcoal 13 skaia / bedrock Sl. 2.86: Prostor med osrednjo in južno cerkvijo. Jugovzhodni sektor, planum 6 - detajl. M. = 1:50. Fig. 2.86: Space between the main and south church, southeastern sector, planum 6 - detail. Scale = 1:50. I 91 siv grušč / grey gravei ■ 107 rjava prst/brown soil ■ 108 rjava glina / brown clay ■ 109 rdeča glina / red clay [H skala/bedrock Sl. 2.87: Prostor med osrednjo in južno cerkvijo, severozahodni sektor, planum 6. M. = 1:50. Fig. 2.87: Space between the main and south church, northwestern sector, planum 6. Scale = 1:50. x = 85,00 y = 26,90 z = 407,30 I I 01 ruševina / destruction layer ■ 64 temno rjava prst / dark brown soil ga 69 siv grušč / grey gravel ■ 70 rjava prst / brown soil gS 71 rjav grušč / brown gravel ■ 73 temno rjava prst z ogljem / dark brown soil with charcoal ■ 74 trda maltna plast 5 / hard layer of mortar 5 g 76 rjav grušč / brown gravel □ 88 droben pesek / fine sand [33 90 rjav grušč / brown gravel iS 91 siv grušč / grey gravel ■ 100 temno rjava prst z ogljem / dark brown soil with charcoal ■ 102 rjava prst / brown soil ■ 107 rjava prst / brown soil ■ 108 rjava glina / brown clay [5] opeka / teguale m skala / bedrock Sl. 2.88: Presek 2 v prostoru med osrednjo in južno cerkvijo. M. = 1:50. Fig. 2.88: Section 2 in the space between the main and south church. Scale = 1:50. drobnega rumenkastega peska, prsti in manjših kamnov (sl. 2.81). Nad njo so večino prostora zavzemale plasti malte različne kakovosti in debeline (sl. 2.81, 2.89). Tik nad SE 65 je ležala SE 63, to je plast na površini dokaj trde, v spodnjem delu pa prhke malte. Nad SE 63 je bilo dokumentiranih še nekaj manjših zaplat trde malte (SE 82) ter plast mehke, neravne malte, v kateri j e bilo veliko drobnih do srednj e velikih kamnov in delov strešnikov (SE 84; sl. 2.90). Above SU 63 a few small patches of compact mortar were documented (SU 82) as well as a layer of loose, uneven mortar which included a lot of small to medium sized stones and roof tiles (SU 84; Fig. 2.90). In the part under the 'memoria' the bedrock was covered by layers of sharp gravel (SU 80 and SU 83), while the strip alongside the south church was covered by sandy soil (SU 81). The gravel was covered by a layer of soil, mixed with charcoal (SU 79) which included some animal bones. ■ 63 trda maltna plast 1 / hard layer of mortar 1 ■ 65 rjava peščena prst s kamni / brown sandy soil with stones I I 75 svetlo rjava peščena prst / light brown sandy soil ■ 78 rjava prst s kamni /brown soil with stones ■ 79 temno rjava prst z ogljem / dark brown soil with charcoal [SS 80 rjav grušč / brown gravel ■ 81 sivo-rjava peščena prst / grey-brown sandy soil ■ 82 trda maltna plast 2 / hard layer of mortar 2 ES! 83 siv grušč / grey gravel ■ 84 mehka maltna plast / loose layer of mortar opeka / tegulae ma skala / bedrock Sl. 2.89: Prostor med osrednjo in južno cerkvijo, jugovzhodni sektor, planum 5. M. = 1:50. Fig. 2.89: Space between the main and south church, southeastern sector, planum 5. Scale = 1:50. Sl. 2.90: Prostor med osrednjo in južno cerkvijo. Jugovzhodni sektor na nivoju maltnih plasti. V sredini očiščen recentni vkop. Fig. 2.90: Space between the main and south church, southeastern sector, mortar layers. The recent trench in the middle. [go 76 rjav grušč / brown gravel ■ 100 temno rjava prst z ogljem / dark brown soli with charcoal ■ 102 rjava prst / brown soil r*1 opeka / tegulae g skala / bedrock x = 88 y = 28 Sl. 2.91: prostor med osrednjo in južno cerkvijo, severozahodni sektor, planum 4. M. = 1:50. Fig. 2.91: Space between the main and south church, northwestern sector, planum 4. Scale = 1:50. S 407 41 74 trda maltna plast 5 / hard layer of mortar 5 76 rjav grušč / brown gravel ■ 100 temnorjava prst z ogljem/ dark brown soil with charcoal ■ 101 Siva prst s kamni / grey soil with stones ■ 102 rjava prst / brown soil ■ 103 pesek, malta In kamni / sand, mortar and stones opeka / tegulae skala / bedrock (I) ■ 63 trda maltna plast 1 / hard layer of mortar 1 ■ 64 temno rjava prst / dark brown soil ■ 65 rjava peščena prst s kamni / brown sandy soil with stones I I 75 svetlo rjava peščena prst / light brown sandy soil ■ 78 rjava prst s kamni /brown soil with stones ■ 79 temno rjava prst z ogljem / dark brown soil with charcoal ■ 81 sivo-rjava peščena prst / grey-brown sandy soli S] opeka / fegu/ae iffl skala/bedrock Sl. 2.93: Prostor med osrednjo in južno cerkvijo, jugovzhodni sektor, planum 3. M. = 1:50. Fig. 2.93: Space between the main and south church, southeastern sector, planum 3. Scale = 1:50. V delu pod "memorijo" so skalno osnovo prekrivale plasti ostrorobega grušča (SE 80 in SE 83), pas ob južni cerkvi pa je zavzemala peščena prst (SE 81). Nad plastmi grušča je ležala še plast žganine (SE 79), v kateri je bilo najdenih le nekaj živalskih kosti. V severozahodnem delu je bila pod južno cerkvijo intenzivna plast grušča (SE 76), v osrednjem delu nad SE 90 pa plast žganine SE 100 in rjava prst SE 102 (sl. 2.88, 2.91). Nad SE 100 in 102 so ležale različne maltne plasti (sl. 2.88, 2.92). Nad SE 100 in SE 102 je tako ležala In the northwest part of the area the space under the south church was covered by a thick layer of gravel (SU 76), while the central part above SU 90 was covered by a layer of soil, mixed with charcoal (SU 100) and brown soil (SU 102; Figs. 2.88, 2.91). Above SU 100 and 102 various mortar layers started to appear (Figs. 2.88, 2.92). SU 100 and SU 102 were thus covered by a layer of loose sand, stones and large chunks of mortar (SU 103), which was in turn covered by a layer of a more compact mortar (SU 74) that covered a part under the main church. This is an Sl. 2.92: Prostor med osrednjo in južno cerkvijo, severozahodni sektor, planum 3. M. = 1:50. Fig. 2.92: Space between the main and south church, northwestern sector, planum 3. Scale = 1:50. iZg 71 rjav grušč / brown gravel ■ 73 temno rjava prst z ogljem / dark brown soil with charcoal ■ 74 trda maltna plast 5 / hard layer of mortar 5 I I 75 svetlo rjava peščena prst / light brown sandy soil Sl 76 rjav grušč / brown gravel ■ 100 temno rjava prst z ogljem / dark brown soil with charcoal ■ 101 Siva prst s kamni / grey soil with stones ■ 102 rjava prst / brown soil ■ 103 pesek, malta in kamni / sand, mortar and stones S] opeka / tegulae H] skala/bedrock Sl. 2.94: Prostor med osrednjo in južno cerkvijo, severozahodni sektor, planum 2. M. =1:50. Fig. 2.94: Space between the main and south church, northwestern sector, planum 2. Scale = 1:50. mehka plast peska, kamnov in večjih kosov malte (SE 103), nad njo pa še trdnejša maltna plast SE 74 v delu pod osrednjo cerkvijo. Gre za približno 15 cm debelo plast, ki je bila na vrhu trda, zglajena in je spominjala na estrih. Obe plasti (SE 103 in 74) močno padata od severozahoda proti jugovzhodu (proti notranjosti izkopnega polja). V dolžini treh metrov padeta za pribl. 0,5 m. SE 103 je v sredini prebil vkop nepravilne okrogle oblike (SE 101). V večini jugovzhodnega sektorja je maltne plasti SE 63 in SE 82 prekrivala do 40 cm debela, mehka plast manjših kamnov in prsti SE 64 (sl. 2.81, 2.93), v njej so bile tudi večje zaplate malte ter veliko strešne opeke. Plast je bila močno kulturna. Tudi v severozahodnem delu prostora so bile plasti, ki so pokrivale trdo maltno plast SE 74, močno kulturne. Ob robu izkopa, ob liniji preseka 2, je maltni plasti approximately 15 cm thick layer that was compact and levelled out on top, and looked like a mortar floor. Both layers (SU 103 and SU 74) sharply (by about 0.5 m over a distance of 3 m) descend from the northwest to the southeast (towards the interior of the excavation area). SU 103 was in the centre punctured by an irregularly round shaped pit (SU 101). In most of the southeast sector the mortar layers SU 63 and SU 82 were covered by a 40 cm thick, soft layer of smaller stones and soil SU 64 (Figs. 2.81, 2.93), which included also large patches of mortar and large quantities of roof tiles. The layer was full of manmade remains. In the northwest sector the layers that covered the hard mortar layer SU 74 were also full of manmade finds. At the edge of the excavation area, alongside the line of section 2, the mortar layers SU 74 and SU 103 ■ 64 temno rjava prst / dark brown soil HSi 69 siv grušč / grey gravel Bi 71 rjav grušč / brown gravel ■ 74 trda maltna plast 5 / hard layer of mortar 5 ■ 93 pesek in malta / sand and mortar I I 94 rjava zemlja / brown soil ■ 96 črna prst z ogljem / black soil with charcoal I I 97 Siva prst / grey soil HI opeka / tegulae S] skala / bedrock Sl. 2.95: Prostor med osrednjo in južno cerkvijo, severozahodni sektor, planum 1. M. = 1:50. Fig. 2.95: Space between the main and south church, northwestern sector, planum 1. Scale = 1:50. SE 74 in SE 103 prekrivala do 30 cm debela plast prsti, mešane z žganino SE 73, v osrednjem delu pa smo še vedno sledili vkopu SE 101 (sl. 2.94). v osrednjem delu in ob severovzhodni stranici je ležala, podobno kot tudi v jugovzhodnem delu prostora, močno kulturna, do 40 cm debela SE 64 (sl. 2.95). Ob skali, na kateri je stala južna cerkev, so nad žganino SE 30 ležale močne erozijske plasti ostrorobega grušča (SE 69, 71). Na severozahodnem robu, pod narteksom osrednje cerkve, pa je nad SE 64 ležala še ena plast bolj mehke malte SE 93, v kateri je bilo tudi nekaj večjih kosov ometa in veliko strešnikov (sl. 2.95). v SE 64 je bil vkopan skeletni grob 21 (SE 99). vkopan je bil plitvo, brez izrazite grobne jame. Glava, usmerjena proti zahodu, je ležala pribl. 30 cm nižje kot noge (sl. 2.95; za grobne pridatke glej: Tonovcov grad. Najdbe, pogl. 2.3). were covered by an up to 30 cm thick layer of soil mixed with charcoal (SU 73), and in the central part we could still follow the pit SU 101 (Fig. 2.94). Similar to the southeast sector the central and northeast sector also revealed a culturally rich layer (SU 64) up to 40 cm thick (Fig. 2.95). At the rock upon which the south church stood the burnt matter SU 30 is covered by strong erosion layers of sharp gravel (SU 69, 71). On the northwest edge, under the narthex of the main church, and above SU 64 lay another layer of loose mortar SU 93, which included large pieces of plaster and large quantities of roof tiles (Fig. 2.95). The skeletal grave 21 (SU 99) was cut into SU 64 (Fig. 2.95). The grave was a shallow one, and did not have an explicit grave pit. The head, oriented towards the west, lay approx. 30 cm lower than the legs (for grave goods see Tonovcov grad. Finds, chapter 2.3). Nad SE 64 je ležala še plast velikih kamnov in SU 64 was covered by a layer of large stones and hu- humusa (sl. 2.81). Plast v preseku 2 ni vidna, ker je bila mus (Fig. 2.81). The layer cannot be seen in the section 2, odstranjena že ob izkopavanju cerkva leta 1996. because it was removed during the 1996 excavations. 2.6 cisterna (vodni zbiralnik) 2.6 water cistern (reservoir) Na najvišjem, vzhodnem platoju Tonovcovega gradu (412 m n. m.), so bili raziskani ostanki vodnega zbiralnika, ki so bili pred začetkom izkopavanj v reliefu videti kot obzidan četverokoten vglobljen prostor. njegov južni del je bil zapolnjen z velikimi kamni (sl. 2.96). objekt je bil delno poškodovan med 2. svetovno vojno, ko je nemška vojska na tem mestu kopala strelske jarke. Pri tem so en jarek izkopali prečno čez južni del objekta. Jarek je bil sicer delno zasut, a še vedno jasno viden. The remains of a large water cistern were investigated on the highest, eastern plateau of tonovcov grad (412 a.s.l.). Prior to the excavations, the remains of the structure were visible as a walled-in, square and deepened area. Its southern section was filled with large stones (Fig. 2.96). The structure was partially damaged during the time of World War II, when the German army dug trenches here and broke through the southern part of the structure. The trench had been partly filled back, but is still clearly visible. Sl. 2.96: Cisterna pred izkopavanjem Pogled s severa. Fig. 2.96: Cistern before excavations. A view from the north. območje je bilo raziskano v dveh akcijah. Leta 2002 je bila odstranjena plast ruševine po zidovih ter očiščen in poglobljen jarek preko južnega dela (presek 1, sl. 2.97,2.98). v notranjosti je bila narejena manjša sonda, s katero je bil ugotovljen nivo dna cisterne. okrog 40 cm prsti in kamnov na dnu objekta je ostalo kot zaščita maltnega estriha.2 The area was researched in two campaigns. In 2002, the destruction layer that covered the walls, was removed, and the trench in the south part was cleaned and deepened (section 1, Figs. 2.97,2.98). A small probe trench was made in the interior of the structure so as to establish the depth of the cistern. approx. 0.4 m of earth and stones in the interior remain as a protection of the mortar floor.2 2 The preliminary report has already been published (Modrijan 2005). 783 ZUNANJI OBODNI ZID / OUTEFI CIRCLE WALL (1) 786 836 886 936 986 4m Sl. 2.97: Shematični tloris cisterne. M. = 1:200. Fig. 2.97: Schematic ground plan of the cistern. Scale = 1:200. Leta 2005 je bil odstranjen še ostanek ruševine v notranjosti cisterne, tako da je bilo doseženo dno po celem objektu. Raziskani objekt je sestavljen iz dveh delov (pril. 5; sl. 2.97). Prvega predstavlja sama cisterna, katere obod je zgrajen z masivnimi dvojnimi zidovi. Zunanji obodni zid je širok približno 40 cm, nanj pa je bil z notranje strani z veliko količino malte prizidan približno 60 cm debel notranji obodni zid. Drugi del predstavlja ograja, ki je potekala okrog cisterne. Med cisterno in ograjo je bil približno 60-80 cm širok pas rjave prsti SE 202 (pril. 5). 2.6.1 STRATIGRAFIJA Plasti in zidovi ter stratigrafski odnosi med njimi so najbolje razvidni iz preseka 1 v očiščenem nekdanjem strelskem jarku, ki je prebil tako južni dvojni obodni zid (zidova 216 in 212), kot tudi južni zid ograje (SE 208; sl. 2.98, 2.99). Jarek je bil približno do višine 411 m zapolnjen z ruševinskimi kamni in zemljo (SE 219). Prvotni vkop za jarek je segal precej globoko, saj smo na prve intaktne plasti naleteli šele na globini približno 410,4 m, pa še to samo v južnem delu jarka, medtem ko In 2005, the remains of destruction layer were removed and the bottom of the whole structure was reached. The researched structure was made from two parts (Insert 5; Fig. 2.97). The first was represented by the water cistern itself, composed of two adjoining walls; the outer circle wall measured approx. 40 cm thick and the inner circle wall, measuring approx. 60 cm thick, was added with a large quantity of mortar to the inside of it. The second part of the structure represented a fence around the cistern. Between the walls of the cistern and the fence was a layer of brown soil (SU 202) measuring approximately 60-80 cm thick (Insert 5). 2.6.1 STRATIGRAPHY The layers, walls and stratigraphic relationships among them are best seen in the section 1 in the former trench, which broke through the southern double circle of the cistern (walls 216 and 212) and also through the southern wall of the fence (SU 208; Figs. 2.98,2.99). The trench was filled with ruination stones and earth (SU 219) approximately to a height of 411 m. The first intact B ^1,00 IH 201 humus z ruševino / destruction iayer 202 rjava prst / brown soii 218 rdeče-rjava mehl \"0,53yjf x = 134 y = 52 x=136 y = 52 ■ 223 črna prst s kamni / black soli with stones I I 226 siva trda prst / grey compact soii ■ 233 rjava prst / brown soii m skaia / bedrock Sl. 2.102: Cisterna, planum 3 v recentnem jarku. M: = 1:50. Fig. 2.102: Cistern, planum 3 in the recent trench. Scale = 1:50. segal do 411,2 m n. m. Na tem mestu neposredno iz ometa prehaja tanka, trda plast malte SE 222, ki najprej nekoliko pade, nato pa na globini 410,90 m poteka horizontalno do skalnega roba. Glede na potek plasti in na dejstvo, da zid od tu navzdol ni več ometan, domnevamo, da predstavlja spodnjo hodno površino. Približno 20 cm višje poteka še ena tanka plast malte SE 220, ki jo od SE 222 ločuje samo pas oranžne gline SE 221. Nad zgornjo maltno plastjo SE 220 leži še plast rdeče-rjave, mehke prsti SE 218, v kateri so bili drobci oglja. Plast je dokaj debela ob zidu 212, kjer sega do vrha še ohranjenega zidu, nato pa njen nivo pade. SE 218 prekriva še pribl. 20 cm debela plast trde, rjave prsti (SE 237), v katero je bil vkopan zid ograje SE 208. Zid je bil sicer zidan z malto, ki pa se je ohranila samo v notranjosti zidu in je bila slabše kvalitete kot tista iz zidov 212 in 216. Dno je bilo na gl. 409,35 m n.m, kar je pribl. 2,3 m pod današnjo hodno površino okoli cisterne. Delno ga je predstavljala skala, delno pa plast oranžne gline SE 230. Skala je bila premazana s 3-5 cm debelim, grobim estrihom, ki je vseboval tudi dokaj velike koščke zdrobljene opeke, nato pa še s pribl. 3 cm debelo plastjo bolj Sl. 2.103: Dno cisterne - detajl (foto: Ž. Cimprič). Fig. 2.103: Bottom of the cistern - detail (photo: Ž. Cimprič). an approx. 1 cm thick layer of plaster on its outer side to a level of 411.2 m. A thin, hard layer of mortar (SU 222) comes directly out of the plaster at this point; at first it declines somewhat and then at a depth of 410.90 m it runs horizontal as far as to the rock edge. The course of the layers, and the fact that the wall is no longer plastered from here on, leads us to presume that this served as the walking surface. Another thin layer of mortar (SU 220) lies approximately 20 cm above; only a layer of orange soil (SU 221) separates it from SU 222. A reddish-brown, loose soil with charcoal (SU 218) covered the upper mortar layer (SU 220). The layer was quite thick along wall SU 212, where it extended to the top of the still preserved wall, then the level drops. The wall of the fence (SU 208) was dug into a layer of hard brown soil (SU 237) which covered SU 218. The mortar was preserved only within the interior of wall 208 and was of lesser quality than the mortar from walls 212 and 216. the bottom of the cistern was at a depth of 409.35 m a.s.l., which is approximately 2.3 m below the present-day walking surface around the cistern. It was represented partially by rock and partially by a layer of orange clay (SU 230). A 3-5 cm thick layer of coarse red mortar pavement (SU 231) - which also incorporated crumbled, relatively large pieces of brick - covered the rocky bottom of the cistern (Insert 5; Figs. 2.103,2.104). The remains of a thin, relatively fine mortar pavement (SU 238) were also found on top of this coarser mortar floor. Wherever the base was represented by clay, a foundation of pebble-stones was made (SU 242). The pebble-stones were first covered with coarse mortar (SU 231), and over this a covering of a finer grained mortar (SU 238) was made (Figs. 2.103-2.106). Because of a large quantity of added crushed bricks the mortar was of an orange-red colour. A layer of mortar (approx. 3 cm thick) passed on from the floor directly on to the walls (Fig. 2.105). It was preserved approx. 0.5 m above the bottom of the cistern, Sl. 2.104: Pogled na notranjost cisterne z juga (foto: Ž. Cimprič). Fig. 2.104: A view of the interior of the water cistern from the south (photo: Ž. Cimprič). finega maltnega estriha (pril. 5; sl. 2.103, 2.104). Na mestih, kjer je dno predstavljala glinena osnova, pa je bila narejena podlaga iz prodnikov (sE 242), čez katero je bila najprej premazana do 5 cm debela plast grobega (sE 231), nato pa še pribl. 3 cm debela plast finega estriha (sE 238). Malta je bila zaradi precejšnje količine primešane zdrobljene opeke oranžno-rdeče barve (sl. 2.103-2.106). okrog 3 cm debela plast malte je neposredno iz tal prehajala tudi na zidove (sl. 2.105). ohranjena je bila še približno do višine 0,5 m nad dnom cisterne, razen v severozahodnem vogalu (stik zidov sE 213 in sE 214), kjer so z recentnim vkopom (verjetno prav tako iz 2. except in the northwest corner (walls sU 213 and 214), where a recent trench (possibly also from World War II) cut through the plaster and removed the inner line of the circle wall sE 214 (Fig. 2.106). The bottom was covered by an intensively orange layer of up to 10 cm thick plaster that had fallen off from the walls (sU 240). A layer of charcoal (sU 241) lay over the fallen off plaster near the southern inner wall (su 216). The cistern was filled up with a layer of large stones and humus (sU 201, Fig. 2.97). Sl. 2.105: Dno in zid cisterne - detajl (foto: Ž. Cimprič). Fig. 2.105: Bottom and wall of the cistern - detail (photo: Ž. Cimprič). Sl. 2.106: Cisterna, uničen del zidu 213 (foto: Ž. Cimprič). Fig. 2.106: Cistern, demolished part of the wall 213 (photo: Ž. Cimprič). svetovne vojne) prebili omet in odstranili tudi notranjo linijo kamnov obodnega zidu SE 214 (sl. 2.106). Dno je skoraj po celi površini prekrivala do 10 cm debela plast s sten odpadlega ometa intenzivno oranžne barve (SE 240). Tik ob južnem notranjem obodnem zidu (SE 216) pa je tik nad plastjo ometa ležala še intenzivna žganinska plast SE 241. Nad plastjo ometa je bila cisterna skoraj do vrha zatrpana z velikimi kamni in humusom (SE 201, sl. 2.97). 2.6.2 WALLS As mentioned above, the cistern consisted of two adjoining walls, surrounded with another wall (a fence) of lesser quality. The space between the cistern and the fence measured approximately 60-80 cm (Insert 5). INNER CIRCLE WALL (walls SU 213, 214, 215, 216) 2.6.2 ZIDOVI Kot je bilo že zgoraj omenjeno, je cisterno sestavljal dvojni, med seboj povezan venec obodnih zidov, ki ga obdajal še en, slabše grajen venec zidov (ograja), med cisterno in ograjo pa je bilo približno 60 do 80 cm prostora (pril. 5). NOTRANJI OBODNI ZID (zidovi SE 213, 214, 215, 216) Notranji obodni zid je bil večinoma vkopan v glineno osnovo (sl. 2.98). Na nekaterih mestih, kjer je bila skalna osnova višja, je bil temeljen na skalo, npr. zahodni notranji obodni zid SE 213. Notranja stena cisterne je bila izdelana iz lepo zloženih, na zunanji strani obdelanih kamnov ter zamazana z vodotesnim ometom, debelim do nekaj centimetrov. Omet je bil izdelan iz malte, mešane z veliko zdrobljene opeke in je bil rdečkaste barve (sl. 2.103-2.105). Zunanja stran notranjega obodnega zidu, kateri se tesno prilega zunanji obodni zid, je bila izdelana slabše, iz nepravilno postavljenih manjših kamnov, ki so bili med seboj Almost all the inner circle wall was dug into a clay base (Fig. 2.98). In some areas, where the bedrock was at its highest - such as the western inner wall (SU 213) -the foundation was bedrock. All four inner walls had interior sides faced with nicely laid stones, all of which were hewn on the outside and then covered with a few cm of water-resistant mortar. The roughcasting was made of reddish mortar mixed with large quantity of crumbled bricks (Figs. 2.103-2.105). The outer side of the inner circle, upon which the outer circle leans, was of poorer quality. Here there were irregularly placed small stones that were compounded with a large quantity of mortar, and affixed to the outer side with more mortar. OUTER CIRCLE WALL (walls SU 209, 210, 211, 212) The outer circle was dug in less deep than the inner. The souther outer wall (SU 212) reaches to an approximate depth of 410.65 m a.s.l., which is at least 0.60 cm higher than the inner wall 216. It can be seen in the section 1, that the wall here stands with no particular foundation directly on the black layer SU 223 zvezani z zelo veliko malte ter prav tako z veliko malte pritrjeni na zunanji obodni zid. ZUNANJI OBODNI ZID (zidovi SE 209, 210, 211, 212) Zunanji obodni zid je bil vkopan bolj plitvo kot notranji. Južni zunanji obodni zid (SE 212) je segal v globino do približno 410,65 m, kar je na tem mestu vsaj 0,60 cm višje od notranjega obodnega zidu (SE 216). Kot je vidno v preseku, zid tu brez posebnega temeljenja stoji na črni plasti SE 223 (sl. 2.98). Na nekaterih mestih pa zidovi zunanjega oboda stojijo neposredno na živi skali, tako npr. severni zunanji obodni zid SE 210. Tudi zunanji obodni zidovi so neposredno povezani med seboj in so bili zgrajeni hkrati. Kljub temu je opazna različna kvaliteta gradnje. Tako je npr. zid SE 209 zgrajen iz dveh front srednje velikih, dokaj pravilnih kamnov, SE 211 pa je grajen dokaj neenakomerno, na nekaterih mestih iz velikih kamnov, drugje pa iz dokaj majhnih kamnov in drobirja. Obe fronti zunanjega oboda sta izdelani iz pravilnih, lepo zloženih kamnov. V preseku 1 smo lahko ugotovili, da je zunanja stena zidu SE 212 ometana od vrha do višine 411,2 m n. m. Na tej višini omet prehaja v horizontalno plast malte (SE 222), ki je zapolnjevala prostor med zunanjo steno cisterne in ograjo in je verjetno predstavljala hodno površino (sl. 2.98). ZUNANJI ZID - OGRAJA (zidovi SE 205, 206, 207, 208) Ograja nima neposredne povezave z zidovi same cisterne. Od njih se razlikuje tudi po načinu gradnje, saj je bila veliko slabše zgrajena. Zidana je bila iz dveh linij precej velikih kamnov, ki so bili med seboj sicer vezani z malto, ki pa je bila slabše kakovosti kot tista na obodnih zidovih cisterne in se na vrhu ograje sploh ni ohranila. Spodnja linija južnega zidu ograje (SE 208) poteka na višini 411,25 m, medtem ko je najvišja ohranjena točka na zidu SE 208 na višini 412,4 m. Zahodni zid ograje (SE 205) je bil na sredini prekinjen s približno 1,2 m široko odprtino - vhodom v vmesni prostor med ograjo in cisterno (pril. 5). (Fig. 2.98). In some places the walls also stand directly upon the bedrock; such is the case with the northern outer wall (SU 210). The outer walls are all connected and built at the same time. However the difference in quality of masonry is noticeable. Wall SU 209 was built from two lines of medium-sized, regular stones; on the contrary, wall SU 211 was built irregular - on certain spots from large stones and on others from small stones and gravel. The outer walls have both sides faced with regular, nicely laid stones. From section 1 we were able to determine that the outer side of wall SU 212 was plastered from the top to a height of 411.20 m a.s. l. At that height the plaster passed into a horizontal layer of mortar (SU 222) between the outer cistern wall and the fence. It presumably represented the walking surface (Fig. 2.98). OUTER WALL - FENCE (walls SU 205, 206, 207, 208) The fence is not directly connected with the walls of the water cistern. Its building style of lesser quality differs from the building style of the cistern. It was built from two lines of quite large stones, connected with a mortar of lesser quality than the mortar of the inner part. The mortar was not preserved atop the walls. The lower line of the south wall 208 is at 411 m a.s.l. and its highest point at 412.2 m. The western wall of the fence (SU 205) was interrupted in the middle by a wide opening measuring approximately 1.20 m; this opening served as the entryway to the area between the fence and the water cistern (Insert 5). 3. stavbni razvoj in kronologija 3. building development and chronology 3.1 Stavba 1 3.1 Building 1 3.2 Stavbi 2 in 3 3.2 Buildings 2 and 3 3.3 Sklop cerkva 3.3 The ecclesiastical complex 3.4 cisterna (vodni zbiralnik) 3.4 Water cistern (reservoir) 3.1 stavba 1 3.1 building 1 3.1.1 PRAZGODOVINA Prva poselitev na območju, kjer je v pozni antiki stala zidana stavba 1, sega v prazgodovino. Dokumentirana je le s posamičnimi najdbami kamnitih artefaktov v ilovnatih plasteh SE 33 in 39 (glej tudi pogl. 2.3.1 in Tonovcov grad. Najdbe, pogl. 6, sl. 6.1). Plast s prazgodovinskimi ostanki je bila odkrita le v naravno nekoliko vglobljenem severnem in vzhodnem delu izkopnega polja, v kvadrantih 618, 619, 668, 669, 718, 719. Zaradi skromnih najdb ni mogoče natančneje datirati te prve prazgodovinske poselitve na območju stavbe 1, umestimo jo lahko v široko časovno obdobje med paleolitikom in bronasto dobo (glej pogl. 2.2.1 in Tonovcov grad. Najdbe, pogl. 6.5). Na območju izkopnega polja stavbe 1 je bilo najdenih tudi nekaj predmetov, ki sodijo v obdobje starejše in mlajše železne dobe (glej tudi Tonovcov grad. Najdbe, pogl. 6.2, 6.3, sl. 6.2j, vendar so bili vsi najdeni v mlajših plasteh (večina v plasteh druge poznoantične faze). Naselbinske plasti, sočasne s temi predmeti, na območju stavbe 1 niso bile najdene. Kljub temu število in tudi struktura teh najdb dovoljuje domnevo, da je bil hrib obljuden tudi v železni dobi. Verjetno ni šlo za stalno poseljenost, morda lahko domnevamo (podobno kot na bližnjem Gradiču) uporabo dela hriba kot kultno mesto (glej Tonovcov grad. Najdbe, pogl. 6.4). 3.1.2 ANTIKA Nekaj predmetov, najdenih v izkopnem polju stavbe 1, lahko datiramo v čas od 1. st. do sredine 3. st. (glej Tonovcov grad. Najdbe, pogl. 2, t. 1; 7: 1-2; 9: 1-3). Tudi zanje velja, podobno kot za prazgodovinske predmete, da jih ne moremo povezati z nobeno poselitveno fazo, saj jih je bila večina najdena v humusu in v plasteh, datiranih v 6. st. (SE 01, 03, 06). le nekaj jih je bilo najdenih v plasteh, ki pripadajo koncu 4. in prvi polovici 5. st., to je prvi poznoantični fazi (fibula tipa Almgren 68 v SE 30: sl. 3.4: 11 in del noriško-panonskega okova v SE 53: sl. 3.10: 1; glej tudi Tonovcov grad. Najdbe, pogl. 2.1). 3.1.1 PREHISTORY The first settlement in the area where building 1 was erected in Late Antiquity reaches back into prehistoric times. This is documented merely by individual stone artefacts from clay layers SU 33 and 39 (see also chapter 2.3.1 and Tonovcov grad. Finds, chapter 6, Fig. 6.1), which covered the bedrock and the sterile clay layer SU 40. The layer with prehistoric remains was revealed only in the naturally deepened north and east part of the excavation area, in quadrants 618, 619, 668, 669, 718 and 719. Due to the modest remains this first prehistoric settlement in the area ofbuilding 1 cannot be dated more precisely; however it can be placed in the broad time span between the Paleolithic and the Bronze Age (see chapter 2.2.1 and Tonovcov grad. Finds, chapter 6.5). A few objects belonging to the Early and Late Iron Age (see also Tonovcov grad. Finds, chapters 6.2, 6.3, Fig. 6.2) were also discovered in the excavation area of building 1, however they were all found in later layers (most in layers belonging to the Late Antiquity phase 2). No settlement layers that would be contemporary to these objects were discovered in the area of building 1. Regardless of this the number and structure of these finds allow for the assumption that the hill was settled in the Iron Age. It is unlikely that this was a permanent settlement, however it could be assumed that a part of the hill - similar to the nearby Gradič - was used as a cult place (see Tonovcov grad. Finds, chapter 6.4). 3.1.2 ANTIQUITY Some objects found in the excavation area of building 1 can be dated to the period between the 1st and mid 3rd century (see Tonovcov grad. Finds, chapter 2, Pls. 1; 7: 1-2; 9: 1-3). Similar to the prehistoric finds they cannot be linked to any settlement phase, for most of them were found in the humus and in the layers that could be dated into the 6th century (SU 01, 03, 06). Only a few were found in layers that could be dated to the 4th and first half of the 5th centuries, i.e. the late Antiquity Podobno velja za novce. Na območju stavbe 1 je bilo najdenih 8 novcev iz časa 1. in 2. st. (glej Tonovcov grad. Najdbe, pogl. 5.1, kat. št. 1-7, 129).1 Vsi so bili zelo izrabljeni, zato na njihovi podlagi ni mogoče govoriti o stalnem novčnem obtoku v naselbini (glej Tonovcov grad. Najdbe, pogl. 5.2.1). Trije od njih so bili najdeni v plasteh druge poznoantične faze (SE 26, 67), trije pa tudi v zgodnjesrednjeveški plasti (SE 10: sl. 3.18: 6-8). Pri dveh SE ni bila jasna. Prvo, vsaj občasno antično poselitev Tonovcovega gradu lahko zanesljiveje umestimo v drugo polovico 3. st. (glej pogl. 2.2.2), čeprav tudi naselbinski sledovi iz tega časa na območju stavbe 1 niso bili najdeni. Predmeti iz tega obdobja (večinoma novci) so bili najdeni v plasteh kasnejših faz ali v premešanih plasteh. Med novci so prevladovali antoninijani. Na območju stavbe 1 jih je bilo najdenih 9 (glej Tonovcov grad. Najdbe, pogl. 5.1, kat. št. 13-16, 19-21, 24-25). Antoninijani so sploh najbolje zastopani novci na Tonovcovem gradu. Pri tem avtor študije o novcih opozarja na možnost dolge uporabe antoninijanov, vse do začetka 4. st., pa tudi na možnost njihove ponovne uporabe v poznem 4. st. in v začetku 5. st. Tak primer bi lahko predstavljali tudi antoninijani, najdeni v plasteh prve poznoantične faze na območju stavbe 1. Po drugi strani dobra ohranjenost nekaterih primerkov bolj kaže na njihovo kratkotrajno uporabo že v času nastanka ali kmalu po njem (glej Tonovcov grad. Najdbe, pogl. 5.2.1). Med preostalim časovno dobro opredeljivim gradivom lahko v ta čas uvrstimo še nekaj predmetov, najdenih v mlajših plasteh. obročaste fibule, sicer širše datirane v 3. in 4. st. (Tonovcov grad. Najdbe, t. 2: 1-4), so bile najdene v plasteh faz PA 1/2 (SE 24), PA 2 (SE 3, 26) in zgodnjega srednjega veka (SE 10: sl. 3.18: 15), eden izmed dveh ključev v obliki prstana (Tonovcov grad. Najdbe, t. 30: 17) pa v fazi PA 2 (SE 26). Pasna spona oblike D (Tonovcov grad. Najdbe, t. 7: 2) je detektorska najdba pred izkopavanji. Nekaj uvožene keramike, ki bi jo sicer lahko datirali že v 3. st., je bilo najdene v plasteh mlajših faz (SE 04, 10, 23). Gre za afriško sigilato oblike Hayes 58 oz. 32/58 (Tonovcov grad. Najdbe, t. 64: 1-3), ki se je na obravnavanem območju pojavljala vse do konca 4. st. in bi torej lahko sodila tudi v prvo poznoantično fazo (glej Tonovcov grad. Najdbe, pogl. 4.1.1) 3.1.3 POZNA ANTIKA PRVA POZNOANTIČNA FAZA (PA 1) Prva poznoantična poselitvena faza je bila z naselbinskimi kulturnimi plastmi dokumentirana na vzhodnem, severnem in osrednjem delu izkopnega phase 1 (fibula type Almgren 68 in SU 30: Fig. 3.4: 11 and a part of the Norico-Pannonian belt plate in SU 53: Fig. 3.10:1; see also tonovcov grad. Finds, chapter 2.1). Similar holds true for coins. Eight 1st and 2nd century coins were found in the area of building 1 (see tonovcov grad. Finds, chapter 5.1, Cat. Nos. 1-7, 129).1 All of them were heavily used, thus they cannot be used to ascertain a constant coin circulation in the settlement (see tonovcov grad. Finds, chapter 5.2.1). Three of them were discovered in Late Antiquity 2 layers (SU 26, 67), while three were discovered in the Early Medieval layer (SU 10: Fig. 3.18: 6-8). For two coins the SU was not clear. The first, at least occasional Antique settlement on tonovcov grad, can be dated to the second half of the 3rd century (see chapter 2.2.2), even though settlement traces from this period were not found in the area covered by building 1. Objects from this period (mainly coins) were discovered in layers belonging to later phases or in mixed layers. With 9 finds in the area of building 1 antonin-ians are the most common coins from this period (see tonovcov grad. Finds, chapter 5.1, Cat. Nos. 13-16, 19-21, 24-25). In general, antoninians are the best represented coins at tonovcov grad. At this the author of the study on coins draws attention to the possibility that these coins were used as late as the beginning of the 4th century, or perhaps put into circulation again in the late 4th and early 5th century. The possibility of their reintroduction is indicated by the antoninians discovered in the Late Antiquity 1 layers in the area of building 1. on the other hand, the well preserved examples indicate that they were used for a brief period of time after minting or shortly afterwards (see tonovcov grad. Finds, chapter 5.2.1). A few other finds discovered in the later layers can also be reliably dated into this period. The ring fibulae, broadly dated into the 3rd and 4th centuries (tonovcov grad. Finds, Pl. 2: 1-4) were discovered in layers belonging to LA 1/2 (SU 24), LA 2 (SU 3, 26) and the Early Middle Ages (SU 10: Fig. 3.18: 15), and one of the two rings-keys (tonovcov grad. Finds, Pl. 30: 17) was discovered in a layer belonging to phase LA 2 (SU 26). The D-shaped belt buckle (tonovcov grad. Finds, Pl. 7: 2) is a metal detector find from before the excavations. Some fragments of the imported pottery that could otherwise be dated to the 3rd century were found in later layers (SU 04, 10, 23). This is the African Red Slip Ware form Hayes 58 or 32/58 (tonovcov grad. Finds, Pl. 64: 1-3), which can be found in the Southeastern Alps until the end of the 4th century and can thus belong also into the Late Antiquity phase 1 (see tonovcov grad. Finds, chapter 4.1.1). 1 Ostali novci iz tega obdobja (glej Tonovcov grad. Najdbe, pogl. 5.1, kat. št. 8, 130, 131, 132, 133) so detektorske najdbe pred začetkom izkopavanj. 1 The other coins from this period (see chapter 5.1, Cat. Nos. 8, 130, 131, 132, 133) are metal detector finds from prior to the excavations. 568 569 669 • • • • • •• • • • keramika / pottery • drobne najdbe / smail finds Sl. 3.1: Razprostranjenost najdb v plasteh prve poznoantične faze v izkopnem polju stavbe 1. Fig. 3.1: Distribution of finds from the Late Antiquity 1 layers in the excavation area of building 1. polja. Datirana je v drugo polovico 4. in začetek 5. st., njen konec pa ni natančno določljiv. Kulturne plasti prve poznoantične faze so bile najdene pod zidovi in plastmi, ki so pripadali drugi poznoantični fazi (glej tudi pogl. 2.3, tab. 2.1). Že v prvi poznoantični fazi so na Tonovcovem gradu stali zidani objekti, ki pa jih natančno ne moremo rekonstruirati. Najdeni ostanki zidov (zidovi 13, 14 in 15) kažejo, da sta morda na tem območju stali dve stavbi (glej pogl. 2.3.2). Ostanek prve sta zidova 13 in 14, ki oblikujeta vogal (sl. 2.2). Žid 4, ki pripada kasnejši stavbi 1, je uničil zid 13 približno 3 m od njegovega stika z zidom 14. Žid 14 je bil v dolžino ohranjen pribl. 2 m, naprej pa je bil odstranjen. Tlorisa tega starejšega objekta tako ni mogoče določiti, njegova orientacija pa je bila podobna orientaciji kasnejše stavbe 1. Močne kulturne plasti prve poznoantične faze (SE 24, 29a) so bile najdebelejše v kvadrantih 669 in 719, kjer sta še delno ohranjena zidova 13 in 14 (sl. 2.3, 3.1), pa tudi geološka osnova je na tem območju najnižja. objekt je verjetno segal tudi v notranjost kasnejše stavbe 1 (kv. 719), kjer pa so bili njegovi ostanki v veliki meri uničeni ob njeni gradnji. V smeri proti severozahodu je bil verjetno omejen s skalno stopnico v kvadrantu 668 (glej pogl. 2.3.2). V SE 24 so bili kasneje vkopani zidovi stavbe 1, zgornji nivo SE 24 pa predstavlja nivo hodne površine stavbe 1. Ostanki zidov 13 in 14 in njuna okolica so bili prekriti s kulturnimi plastmi druge poznoantične faze. Ruševina, ki bi lahko pripadala tema dvema zidovoma, ni bila ohranjena. Domnevamo, da so bili tako deli zidov kot njihova ruševina uporabljeni kot gradbeni material pri gradnji poznejše stavbe 1. Ostanki starejše poselitve so bili odkriti tudi pred vhodom v stavbo 1 (sl. 2.2). Tu je bil v dolžino pribl. 5 m ohranjen zid (zid 15), ki je ležal delno na skalni osnovi, delno pa na plasti SE 36, ki je segala pod zidom 1 tudi v notranjost kasnejše stavbe 1 (sl. 2.6, 2.7). Verjetno je tudi zid 15 ostanek starejše, zelo uničene stavbe. Kje je bila njena notranjost, na podlagi ohranjenih ostankov ni mogoče z gotovostjo pojasniti, saj so bile kulturne plasti prve poznoantične faze ohranjene tako severno kot južno od zidu 15 (sl. 3.1). Poleg tega je stratigrafijo na tem mestu močno poškodovalo drevo, ki je svojimi koreninami uničilo velik del plasti, tako da odnosi med plastmi nad zidom 15 niso popolnoma jasni. Ker na mestu domnevno starejše stavbe (podobno kot pri zidovih 13 in 14) ni bilo ruševine, lahko sklepamo, da so bili tudi ti kamni porabljeni pri zidavi stavbe 1. prvo poznoantično fazo lahko umestimo v drugo polovico 4. st. in na začetek 5. st. Od značilnih predmetov, najdenih v plasteh te faze, jo v ta čas opredeljujejo predvsem novci in uvožena keramika, medtem ko so ostale najdbe redke (sl. 3.2-3.10). Novci druge polovice 4. in začetka 5. st. so najštevilčnejša skupina novcev na območju stavbe 1, med 3.1.3 LATE ANTIQUITY LATE ANTIQUITY PHASE 1 (LA 1) The first Late Antique settlement phase was documented by cultural layers discovered in the east, north and central parts of the excavation area. These layers were dated into the second half of the 4th and the beginning of the 5th century, however it is impossible to define the precise end of this phase. Cultural layers belonging to Late Antiquity phase 1 were found under the walls and layers that belonged to Late Antiquity phase 2 (see also chapter 2.3, tab. 2.1). Masonry buildings stood on Tonovcov grad already in Late Antiquity phase 1; however, it is impossible to reconstruct them with precision. The discovered wall remains (walls 13, 14 and 15) indicate the possibility of two buildings standing in this area (see chapter 2.3.2). The remains of the first building are represented by walls 13 and 14, which form a corner (Fig. 2.2). Wall 4, which belongs to the later building 1, destroyed wall 13 approximately 3 m from its contact with wall 14. Approximately 2 metres of wall 14 were preserved in length, while the continuation of it was removed. The ground plan of this older building could thus not be ascertained, but its orientation was similar to the orientation of the later building 1. The strong Late Antiquity phase 1 cultural layers (SU 24, 29a) were thickest in quadrants 669 and 719 where walls 13 and 14 are partially preserved (Figs. 2.3, 3.1) and the geologic base is at its lowest. It is highly likely that the building reached into the interior of the later building 1 (qu. 719) where its remains were most likely destroyed during construction. Towards the northwest it most likely reached as far as the rock step in quadrant 668 (see chapter 2.3.2). The walls of building 1 were later on dug into SU 24, at which the upper level of SU 24 represented the floor level of building 1. The remains of walls 13 and 14 and their surroundings were covered by cultural layers belonging to Late Antique phase 2. The ruins that could have belonged to these two walls were not preserved. We can assume that parts of the walls as well as their ruins were used as building material in the construction of the later building 1. The remains of an earlier settlement were discovered also in front of the entrance into the building 1 (Fig. 2.2). In this area approximately 5 metres of wall 15 was preserved in length. The wall stood partially on the bedrock and partially on layer SU 36, which ran under wall 1 into the interior of the later building 1 (Figs. 2.6, 2.7). The wall 15 is probably also a remnant of an older, severely destroyed building. The location of its interior cannot be explained by the preserved remains, as Late Antiquity phase 1 cultural layers were preserved to the north as well as to the south of wall 15 (Fig. 3.1). An additional problem is represented by the fact that the stratigraphy in this area Sl. 3.2: Najdbe iz faze PA 1 (SE 04). 1 novec; 2-5 železo; 6-19 keramika. M. 2-5 = 1:2; 6-19 = 1:3. Fig. 3.2: Finds from LA 1 phase (SU 04). 1 coin; 2-5 iron; 6-19 pottery. Scale 2-5 = 1:2; 6-19 = 1:3. SE / SU 29A nljCCUJ 10 J 12 ........ ..... 13 F I—< 14 t \ 15 16 1 Sl. 3.3: Najdbe iz faze PA 1 (SE 29a). 1,2,4-6 železo; 3 rogovina; 7,8 steklo; 9-17 keramika. M. 1-9 = 1:2; 10-17 = 1:3. Fig. 3.3: Finds from LA 1 phase (SU 29a). 1,2,4-6 iron; 3 antler; 7,8 glass; 9-17 pottery. Scale 1-9 = 1:2; 10-17 = 1:3. Sl. 3.4: Najdbe iz faze PA 1 (SE 29a in 30). 1-8,21 keramika; 9,10 novca; 11,13 bron, 12,14-20 železo. M. 1-8,21 = 1:3; 11-20 = 1:2. Fig. 3.4: Finds from LA 1 phase (SU 29a and 30). 1-8,21 pottery; 9,10 coins; 11,13 bronze; 12, 14-20 iron. Scale 1-8,21 = 1:3; 11-20 = 1:2. - Sl. 3.7: Najdbe iz faze PA 1 (SE 17, 31, 32, 36). 1,5 železo; 2 bron; 3,8-10 novci; 4 svinec; 6-7 keramika; 11-14 rogovina in železo. M. 1-2,5,11-14 = 1:2; 4 = 2:1; 6-7 = 1:3. Fig. 3.7: Finds from LA 1 phase (SU 17, 31, 32, 36). 1,5 iron; 2 bronze; 3,8-10 coins; 4 lead; 6-7 pottery; 11-14 antler and iron. Scale 1-2,5,11-14 = 1:2; 4 = 2:1; 6-7 = 1:3. Sl. 3.8: Najdbe iz faze PA 1 (SE 36). 1-3,5-9 železo; 4 svinec; 10 steklo; 11-22 keramika. M. 1-10 = 1:2; 11-22 = 1:3. Fig. 3.8: Finds from LA 1 phase (SU 36). 1-3,5-9: iron; 4 lead; 10 glass; 11-22 pottery. Scale 1-10 = 1:2; 11-22 = 1:3. SE / SU 36 II f \\ 12 14 15 16 17 Sl. 3.10: Najdbe iz faze PA 1 (SE 53, 57, 60, 77, 78). 1 bron; 2 železo; 3 novec; 4-10 keramika. M. 1,2,7 = 1:2; 4-6,8-10 = 1:3. Fig. 3.10: Finds from LA 1 phase (SU 53, 57, 60, 77, 78). 1 bronze; 2 iron; 3 coin; 4-10 pottery. Scale 1,2,7 = 1:2; 4-6,8-10 = 1:3. njimi pa so najbolje zastopani novci kovnega obdobja 408-423, kovani v času samostojne vlade cesarja Honorija (glej Tonovcov grad. Najdbe, pogl. 5.2.1, kat. št. 91-104). Večina novcev druge polovice 4. in začetka 5. st. je bila sicer najdena v plasteh druge poznoantične faze, 6 novcev pa tudi v plasteh prve poznoantične faze (SE 04, 30, 68). Med keramiko je dobro časovno opredeljivo uvoženo gradivo (afriška sigilata, oljenke, amfore, glazirana keramika; glej Tonovcov grad. Najdbe, pogl. 4.1). Med namiznim posodjem prevladuje tako imenovana navadna namizna keramika (Tonovcov grad. Najdbe, t. 65: 11,12; 66). Sigilatni krožniki in sklede (Tonovcov grad. Najdbe, t. 64: 1-9, 11; 65: 1,2) so razmeroma slabo zastopani, skromen je tudi nabor oblik. Dokaj pogoste in tudi tipološko raznolike pa so amfore (Tonovcov grad. Najdbe, t. 67; 68; 69: 1-3), kar kaže na živahne stike s sredozemskim prostorom (glej Tonovcov grad. Najdbe, pogl. 4.1). was severely damaged by a tree that ruined a large part of the layers, thus the relations between the layers above wall 15 are not entirely clear. Because no ruins were discovered at the location of the assumed older building (similar as with walls 13 and 14), we can assume that the stones were used for the construction of building 1. late Antiquity phase 1 can be dated to the second half of the 4th century and the beginning of the 5th. Coins and imported pottery are the most typical objects found in the layers belonging to this phase, while other finds are rare (Figs. 3.2-3.10). Most of the coins discovered in the area of building 1 can be dated to the second half of the 4th and beginning of the 5th century, and within this group most coins were minted between 408 and 423, in the time of the independent rule of Emperor Honorius (see Tonovcov grad. Finds, chapter 5.2.1, Cat. Nos. 91-104). Most of the coins from the second half of the 4th and beginning of omenjeno posodje večinoma sodi v čas druge polovice 4. in prve polovice 5. st. Nekaj keramičnih predmetov (npr. afriška sigilata oblik Hayes 32/ 58 in 58) je sicer mogoče datirati že v pozno tretje in zgodnje 4. st., čeprav so bili na območju jugovzhodnih Alp v uporabi še do konca 4. st. ostale oblike kažejo predvsem na konec 4. in prvo polovico 5. st. (glej Tonovcov grad. Najdbe, pogl. 4.1.1). Posode so bile najdene tako v plasteh prve poznoantične faze (sl. 3.2: 6-9; 3.3: 10-12; 3.4: 21; 3.5; 3.8: 11; 3.10: 4) kot v mlajših plasteh. V nekaj primerih je bil del posode najden v plasti PA 1, njeni ostali deli pa v ruševinskih plasteh. Tudi to kaže na veliko uničenje plasti prve poznoantične faze, ki ga je povzročila gradnja stavbe 1.2 Nekeramične najdbe iz plasti prve poznoantične faze so redke in neizpovedne, lahko rečemo, da izražajo bivalni značaj objekta (deli nakita, glavnikov, stavbne opreme, orodja: sl. 3.2: 2-5; 3.3: 1-6; 3.4: 12-20; 3.7: 1,2,5,11-14; 3.8: 1-10; 3.10: 2). Prav tako so zelo redke najdbe stekla (sl. 3.3: 7,8; 3.8: 10; Tonovcov grad. Najdbe, t. 57: 1,15-16). Veliko kovinskih predmetov, ki sodijo v drugo polovico 4. in prvo polovico 5. st., pa je bilo najdenih tudi v plasteh mlajše, to je druge poznoantične faze (sl. 3.11). V njih je bilo odkritih več delov poznorimskih pasnih garnitur, fibul (Tonovcov grad. Najdbe, t. 2: 8,11-12; 7: 4-5) in novcev iz konca 4. st. in prvih desetletij 5. st., ki kažejo na možno prisotnost vojske, med njimi tudi visokih vojaških predstavnikov (glej Tonovcov grad. Najdbe, pogl. 2.1.3). DRUGA POZNOANTIČNA FAZA (PA 2) V drugi poznoantični fazi je bila postavljena stanovanjska stavba 1 s prizidkom (glej pogl. 2.3.3). Kot je razvidno že iz opisa zidov, so bili najprej hkrati postavljeni zidovi osrednjega prostora, pravokotno nanj pa prizidana zidova vetrolova. Tudi stranski prostor je bil prizidan h glavnemu. Vsi notranji zidovi glavne stavbe in večina zunanjih so bili ometani z malto (glej pogl. 2.3.3). Plast s sten odpadlega ometa ob zidovih je bila debela do 0,5 m. Izjema je le zunanja stran zidu 3, obrnjena proti skalnemu pobočju, ki ni bila ometana. Prav tako niso bili ometani zidovi prizidka. Zidovi stavbe 1 so bili deloma postavljeni na skalno osnovo, deloma pa vkopani v glineno osnovo (SE 40) oziroma v različne starejše kulturne plasti (SE 04, 24, 29a: sl. 2.3,2.5-2.6,2.8-2.9). Te plasti so bile v okolici stavbe debele do pol metra, medtem ko so bile v notranjosti ohranjene le v skromnih ostankih (SE 17, 20, 36, 77, 78: sl. 2.5,2.9,2.14). Pri gradnji stavbe 1 je bil očitno uporabljen gradbeni material starejših faz poselitve, predvsem prve poznoantične faze (glej zgoraj). the 5th century were discovered in Late Antiquity phase 2 layers, while 6 coins were discovered in Late Antiquity phase 1 layers (SU 04, 30, 68). The imported pottery can be precisely dated (African Red Slip Ware, oil lamps, amphorae, glazed pottery; see tonovcov grad. Finds, chapter 4.1). The so-called coarse tableware dominates within the group of tableware (tonovcov grad. Finds, Pls. 65: 11,12; 66). Sigillata plates and bowls (tonovcov grad. Finds, Pls. 64: 1-9,11; 65: 1,2) are relatively poorly represented and they come in a limited selection of shapes. Relatively common and also typologically diversified are amphorae (tonovcov grad. Finds, Pls. 67; 68; 69: 1-3), which indicates a lively trade with the Mediterranean (see tonovcov grad. Finds, chapter 4.1). Most of the mentioned vessels can be dated into the second half of the 4th and beginning of the 5th century. Some of the pottery (African Red Slip Ware forms Hayes 32/58 and 58) can be dated into the late 3rd and early 4th century, however in the area of the Southeastern Alps this form remained in use to the end of the 4th century (see tonovcov grad. Finds, chapter 4.1.1). The other forms are mainly indicative of the end of the 4th and the first half of the 5th century. The vessels were found in layers belonging to Late Antiquity phase 1 (Figs. 3.2: 6-9; 3.3: 10-12; 3.4: 21; 3.5; 3.8: 11; 3.10: 4) as well as in later layers. In some cases one part of the vessel was found in LA 1 layer, while the remaining parts of the same vessel were discovered in the destruction layers. This also indicates that the Late Antiquity 1 layers were severely destroyed during the construction of building 1.2 Non-ceramic finds from the Late Antiquity 1 layers are rare and unspecific, they suggest the building was used a living space (parts of jewellery, combs, architectural fittings, tools: Figs. 3.2: 2-5; 3.3: 1-6; 3.4: 12-20; 3.7: 1,2, 5,1114; 3.8: 1-10; 3.10: 2). Glass finds are also very rare (Figs. 3.3: 7,8; 3.8: 10; tonovcov grad. Finds, Pl. 57: 1,15-16). Numerous metal objects that belong into the second half of the 4th and beginning of the 5th century were also found in the layers of the later, Late Antiquity 2 phase (Fig. 3.11). Numerous parts of Late Roman belt sets, fibulae (tonovcov grad. Finds, Pls. 2: 8,11-12; 7: 4-5) and coins indicate that this area was used at the end of the 4th and beginning of the 5th century; they also indicate the possibility of a military presence, even high military officials (see tonovcov grad. Finds, chapter 2.1.3). LAtE ANtlQUItY PHASE 2 (LA 2) Building 1 with the living quarters and its outhouse were erected during Late Antiquity phase 2 (see chapter 2.3.3). 2 Deli glaziranega lončka (sl. 3.2: 6; Tonovcov grad. Najdbe, t. 65: 8) so bili npr. najdeni daleč narazen v kvadrantih 769/A1, 719/D3, 669A/1 in 766/A4 v plasteh faz PA 1 in PA 2. 2 Parts of the glazed mug (Fig. 3.2: 6; Tonovcov grad. Finds, Pl. 65: 8) were found far apart in quadrants 769/A1, 719/ D3, 669A/1 and 766/A4 and in layers belonging to phases LA 1 in LA 2. Sl. 3.11: Predmeti iz druge polovice 4. in prve polovice 5. st., najdeni v plasteh druge poznoantične faze. 1-13 bron; 14 steklo; 15-18 železo. M. = 1:2. Fig. 3.11: Finds of the second half of the 4th and the first half of the 5th century from Late Antiquity 2 layers. 1-13 bronze; 14 glass; 15-18 iron. Scale = 1:2. Zidovi glavnega prostora in vhod s pragom so bili postavljeni hkrati. Zidova vetrolova sta bila sicer prizida-na na zidova stavbe, vendar je glede na enotno zasnovo in enoten način gradnje verjetno, da sta bila zidana hkrati s stavbo, prav tako tudi ognjišče 1 in podstavek za As can be seen already from the descriptions of the walls, the walls of the main room were the first to be erected (and simultaneously), and the two wind shield walls were added at a right angle. The outhouse was also attached to the main building. All inner sides of the steber (zid 12) na zunanji strani zidu 1. Stranski prostor (prizidek) je bil k zidu 2 verjetno prizidan pozneje, saj se način njegove gradnje precej razlikuje od gradnje glavnega prostora. Zidovi prizidka so bili grajeni skoraj brez uporabe malte, slabše temeljeni in neometani. Zid 8 je bil postavljen na plast (SE 14), ki že pripada drugi poznoantični fazi. Kasnejšo dozidavo potrjuje tudi dejstvo, da je bil prizidan na že ometan zid 2 glavnega prostora. Glede na slabo temeljenje in slabšo zidavo prizidka je bil ta verjetno nižji od glavne stavbe. Mogoče je tudi, da je bil zidan samo njegov spodnji del, zgornji pa lesen. Večino uporabnega prostora v njegovi notranjosti je zavzemala skala, nekaj za bivanje primernega ravnega prostora je bilo le ob vhodu, kjer je bilo v severozahodnem vogalu ognjišče (ognjišče 2), ter ob zidu 2. Ker je bila stavba 1 postavljena na zelo neravno podlago, je bila tudi globina temeljev zidov različna. V notranjosti glavnega prostora so ležale plasti trde, steptane gline (SE 18, 21), v katero so bili bolj ali manj na gosto postavljeni pribl. za pest veliki kamni, med njimi pa kosi opek in zaplate malte. Velik del prostora je zavzemala skala (sl. 2.10,2.29). tako neravnemu terenu se je očitno prilagajal tudi hodni nivo v posameznih delih objekta. Ostanki pravega estriha niso bili najdeni. Zgornji nivo steptane ilovice SE 21 ob vhodu v objekt ustreza nivoju praga. to je tudi nivo, kjer lepo izdelan in ometan zid 1 prehaja v grobo izdelan temelj. Glede na to lahko domnevamo, da je imela SE 21 ob vhodu v objekt funkcijo hodne površine. Približno na sredi prostora se nad tem domnevnim hodnim nivojem dviguje skalna osnova, ki ob zidu 3 seže približno 80 cm nad SE 21 (sl. 3.12). Osrednji del prostora ob zidu 3 je zavzemala skala, ki se je dvigala do 1 m nad domnevnim hodnim nivojem ob zidu 1. Glede na tako neravno površino v notranjosti objekta lahko domnevamo različne nivoje hodne površine. Zid 2 je bil temeljen 30-40 cm višje kot zid 1, ruševinske plasti ob njem pa so ležale 20 cm globlje, kot so segali njegovi temelji (sl. 2.29). Ker ni verjetno, da bi bil hodni nivo pod nivojem temeljev zidu, med zidom 2 in skalo lahko domnevamo nekoliko dvignjen lesen pod, ki je premoščal razliko med njima. tako domnevo potrjuje tudi pribl. 10 cm široka reža, izoblikovana v zidu 2 nad njegovimi temelji (sl. 2.15), ki je bila lahko opora za lesene deske (sl. 3.12). Na ta način bi se nekoliko povečala tudi uporabna površina stavbe, saj bi tak lesen pod prekril tudi del skalne osnove. temelji zidu 4 in vogalnega stika zidov 3 in 4 so precej nižji od temeljev zidov 1 in 2. Na 396,6 m n. m. je bila ob stiku zidov 3 in 4 izdelana nekakšna polička iz malte (sl. 2.18), ki je lahko dajala oporo leseni konstrukciji, morda kotni klopi v bližini ognjišča. Večji lesen pod v tem prostoru ni verjeten, saj bi ognjišče (vrh na nivoju približno 396,5 m n. m.) ležalo pod njim. Zidano odprto ognjišče je ležalo v jugovzhodnem delu prostora (glej pogl. 2.2.3, sl. 2.28-2.29). Delno je bilo postavljeno na skalno osnovo, delno pa na glineno walls of the main room and most of the external ones were covered in plaster (see chapter 2.3.3). The layer of the fallen off plaster alongside the walls was up to 0.5 m thick. The only exception was the external side of wall 3, which was turned toward the rocky slope, for this was not covered in plaster. The walls of the outhouse were also not plastered. The walls of building 1 were partially constructed on the bedrock, and partially dug into the clay base (SU 40) or into various older cultural layers (SU 04, 24, 29a: Figs. 2.3,2.5-2.6, 2.8-2.9). In the vicinity of the building these layers were up to half a metre thick, while in the interior they were only modestly preserved (SU 17, 20, 36, 77, 78: Figs. 2.5,2.9,2.14). It is obvious that the material from the previous settlement phases, especially LA 1 (see above) was used in the construction of building 1. The walls of the main room and the entrance with the threshold were constructed at the same time. The walls of the wind shield were attached to the building walls, but taking into account the unified plan and the unified building style it is likely that the wind shield walls were erected at the same time as the building, as was fireplace 1 and the pedestal for the pillar (wall 12) on the exterior of wall 1. The outhouse was most likely attached to wall 2 later on, for its building style differs greatly from the building style of the main building. Its walls were built almost without any mortar, their foundations were not as strong and they were not covered in plaster. Wall 8 of the outhouse was positioned on top of layer SU 14, which already belongs to Late Antiquity 2. The latter addition is also confirmed by the fact that it was attached to the already plastered wall 2 of the main building. taking into account the poor foundations and the poor construction of the outhouse it is likely that it was not as high as the main building. It is also possible that only the lower part was masonry, while the upper part was wooden. Most of the useable space in its interior was occupied by rock, thus the only levelled out inhabitable space was alongside the entrance - where a fireplace (fireplace 2) was placed in the northwest corner - and along wall 2. Because building 1 was placed on an uneven base, the depths of the wall foundations vary greatly. Layers of hard clay (SU 18, 21) were found in the interior of the main room; within these layers we discovered more or less densely positioned stones (about the size of a fist) with pieces of bricks and patches of mortar scattered amongst them. A large part of the room was occupied by rock (Figs. 2.10,2.29). In certain parts of the building the floor surface was adjusted to this uneven terrain. The remains of a real mortar floor were not discovered. The upper level of the compact clay (SU 21) at the entrance into the building fits the level of the threshold. This is also the level at which the carefully made and plastered wall 1 turns into the rough foundation. taking this into account we can assume that at the entrance into the building SU 21 had the function of the Sl. 3.12: Shematičen presek čez stavbo -poskus rekonstrukcije hodnega nivoja. Fig. 3.12: Schematic section - an attempt at reconstruction of the floor surface level of building 1. Sl. 3.13: Shematičen presek čez stavbo - poskus rekonstrukcije hodnega nivoja. Fig. 3.13: Schematic section - an attempt at reconstruction of the floor surface level of building 1. plast prve poznoantične faze SE 18, domnevno hodno površino v tem delu stavbe. Njegov zgornji nivo je bil približno 15-20 cm nad nivojem SE 18. Glede na količino ruševine, ki je prekrivala ohranjene ostanke zidov stavbe 1, lahko domnevamo, da je bila ta v celoti zidana iz kamna. Verjetno je bila pritlična, zidano ali leseno zgornje nadstropje je malo verjetno zaradi odprtega ognjišča in neurejenega odvoda zraka iz objekta. V ruševinah stavbe je bilo najdene tudi precej strešne opeke, vendar ne dovolj, da bi lahko domnevali, da je bila prekrita z njo. Verjetneje se zdi, da je prišla opeka v ruševino iz sklopa cerkva, ležečih nad stavbo 1. Vhod v stavbo s severozahodne strani je bil zavarovan z vetrolovom. Ostalih stavbnih odprtin ne moremo rekonstruirati. okna glede na zelo skromne ostanke okenskega stekla niso bila zastekljena (sl. 3.14). Če odvajanje dima ni bilo urejeno, potem niti niso smela biti zastekljena (prim. Pleterski 2008b, 27). Vseh ostankov stavbne opreme, najdenih v izkop-nem polju stavbe 1 v plasteh druge poznoantične faze, ne moremo povezati neposredno s stavbo, saj jih je bilo precej tudi v njeni okolici (Tonovcov grad. Najdbe, sl. 2.9). Predvsem predmeti, najdeni v kv. 666 in 619, bi lahko pripadali tudi sosednjemu objektu 26 (sl. 1.7). Kljub temu najdeni deli stavbne opreme omogočajo nekaj sklepov. Vsaj ena vrata (ali skrinja) so bila okovana z masiv-nejšimi in okrašenimi tečaji (Tonovcov grad. Najdbe, t. 33: 1), morda so bila to vhodna vrata. Najdenih je bilo veliko ključev in zapahov, zunanjih okovov ključavnic pa ni bilo. Premična lesena krila vrat ali oken so bila pritrjena s tankimi razcepnimi tečaji (Tonovcov grad. floor surface. Approximately in the centre of the room the bedrock rises above the assumed floor surface and reaches approximately 80 cm above SU 21 (Fig. 3.12). Alongside wall 3 the central part of the room was occupied by bedrock that rose up to 1 metre above the assumed floor level alongside wall 1. Due to such an uneven surface we can assume that various floor levels existed in the building interior. The foundations of wall 2 were between 30 and 40 cm higher than that of wall 1, while the destruction layers alongside it lay 20 cm deeper than its foundations (Fig. 2.29). As it is unlikely that the floor level was below the level of the wall foundations, we assume the existence of a slightly elevated wooden floor between wall 2 and the rock which would overcome the difference in height between the two. Such an assumption is also confirmed by the approximately 10 cm wide fissure in wall 2 (above its foundations, Fig. 2.15) which could have supported wooden planks (Fig. 3.12). This would also slightly increase the useful surface of the building, for such a wooden floor would also cover a part of the bedrock. The foundations of wall 4 and the corner of walls 3 and 4 are quite a bit lower than the foundations of walls 1 and 2. At 396.6 m a.s.l. some sort of a mortar shelf was created in the corner formed by walls 3 and 4 (Fig. 2.18). It can be assumed that it provided support for a wooden construction, maybe a corner bench close to the fireplace. It is unlikely that a wooden floor would be positioned in this area, for the fireplace (peaking at approximately 396.5 m a.s.l.) was immediately below. The open masonry fireplace was located in the southeast part of the room (see chapter 2.2.3, Figs. 2.28- 666 715 • <3> 765 766 815 okensko steklo / window glass 617 618 619 667 668 669 718 N ^20 767 768 A 7"» W&Q Wz 819 820 868 869 917 918 (I) 0 4m Sl. 3.14: Razprostranjenost okenskega stekla v izkopnem polju stavbe 1. Fig. 3.14: Distribution of window glass fragments in the excavation area of building 1. Najdbe, t. 32: 3-24), ki so bili manj močni in vzdržljivi; močnejši so bili morda težje dosegljivi ali zahtevnejši za izdelavo. Veliko žebljev različnih dolžin in nekaj spojk je spenjalo lesene dele hiše in njene opreme (Tonovcov grad. Najdbe, t. 33: 4-9; 34-39). Predmetov, ki bi lahko pomagali rekonstruirati opremo v notranjosti stavbe, ni bilo najdenih veliko. Od delov pohištva so skoraj zagotovo uporabljali skrinje in skrinjice, v katerih so hranili svoje premoženje (glej Tonovcov grad. 2.29). A part of it was placed on the bedrock, while the other part was placed on the clay layer (SU 18) belonging to Late Antiquity phase 1 - assumed to be the floor surface in this part of the building. Its upper part lies approximately 15-20 cm above level SU 18. taking into account the quantity of the ruins that covered the preserved remains of the walls of building 1, we can assume that the building was made entirely from stone. It was probably merely a ground floor building, Najdbe, pogl. 2.1.6; t. 33: 10,11). Ta oblika pohištva se je ohranila od rimske dobe pa skoraj do sedanjega časa. Na mnogih poznoantičnih najdiščih lahko najdemo vsaj dele zapiralnega sistema, če že ne kovinskih okovov. v plasteh druge poznoantične faze so bili najdeni tudi ostanki oljenk (Tonovcov grad. Najdbe, t. 65: 3-6), ker pa gre za oblike, značilne predvsem za 5. st. (glej Tonovcov grad. Najdbe, pogl. 4.1.2), ne moremo z gotovostjo trditi, da so povezane z življenjem hiše, čeprav ta možnost obstaja. Najdeni so bili tudi deli visečih steklenih svetilk (glej Tonovcov grad. Najdbe, pogl. 3.1; t. 57: 19-21). večina vsakdanjih opravil se je verjetno odvijala pred hišo. Ostanek podstavka za steber pred zidom 1 kaže, da je bil prostor pokrit z nadstreškom. Podoben nadstrešek lahko predvidevamo tudi na drugi strani vhoda, ob zidu 5. Ker je na tem mestu skalna osnova precej višja in delno zravnana, je bil steber lahko postavljen neposredno na skalo. Intenzivno dejavnost na prostoru pred hišo potrjujejo tudi bogate kulturne plasti iz druge poznoantične faze ob vhodu v stavbo (predvsem kv. 666, 716, 717 in 766, sl. 3.15). v kvadrantu 666 so bile v kulturnih plasteh druge poznoantične faze najdene skupine ravnih kamnitih plošč, katerih funkcije za sedaj še ne znamo natančneje opredeliti (sl. 2.30). Postavljene so bile v zelo mehke plasti in morda bi lahko šlo za neko vrsto poploščenja pred vhodom v hišo. Ker je bilo nekaj ožganih, smo najprej domnevali, da predstavljajo ognjišče, vendar glede na njihovo veliko gostoto in dejstvo, da jih večina ne kaže nobenih sledov vročine, ta domneva najbrž ni pravilna. Bogate kulturne plasti druge poznoantične faze so ležale tudi na zunanji strani zidu 4. Predvsem je treba omeniti veliko število vzhodnosredozemskih amfor, najdenih ob zunanjem vogalu zidov 3 in 4 (predvsem v kv. 719). Morda lahko tu domnevamo skladiščenje uvoženega blaga (glej Tonovcov grad. Najdbe, pogl. 4.1.8, sl. 4.1). Med keramičnimi najdbami prevladuje groba hišna keramika (lonci in sklede; glej Tonovcov grad. Najdbe, pogl. 4.2). Nekeramično gradivo, najdeno znotraj stavbe 1 in v njeni neposredni okolici, je značilno za samooskrbne staroselske naselbine jugovzhodnoalpskega prostora: kovinski deli hišne opreme, orodje, moška in ženska noša, orožje ter stekleni kozarci na visoki nogi. Prestižnejši uvoženi predmeti ter predmeti germanskega značaja niso pogosti (glej Tonovcov grad. Najdbe, pogl. 2.1 in 3.1). Med novci je v plasteh druge poznoantične faze opazna močna zastopanost (39,7 %) novcev druge polovice 4. in začetka 5. st. (glej Tonovcov grad. Najdbe, pogl. 5.2). Hiša verjetno ni bila porušena naenkrat, ampak je propadala počasi. Intenzivna žganinska plast (SE 11) je prekrivala samo osrednji del glavnega prostora stavbe 1 (sl. 2.9, 2.31), drugje pa žganine, ki bi kazala na večji požar, ni bilo. Na počasno opustitev kaže tudi dejstvo, da je bilo v notranjosti hiše najdenih izredno malo predmetov iz časa njene uporabe (sl. 3.15). for it is unlikely that it had a masonry or wooden upper floor as it had an open fireplace and smoke extraction was not provided. Some roof tiles were discovered within the ruins, however, there were not enough of them for us to assume the building was covered with them. It is more likely that the roof tiles slid down from the ecclesiastical complex above building 1 as these were certainly covered with roof tiles. The entrance into the building from the northwest was protected by a wind shield. The remaining openings cannot be reconstructed. taking into account the extremely rare remains of window glass it is unlikely that the windows were glazed (Fig. 3.14). As the smoke had nowhere else to escape from the building the windows had to be open (cf. Pleterski 2008b, 27). Not all of the remains of the architectural fittings that were discovered in the late Antiquity 2 layers in the excavation area of building 1 can be linked directly to the building, for a lot of them were also found in its surrounding (tonovcov grad. Finds, Fig. 2.9). Especially objects that were found in qus. 666 and 619 could belong to the neighbouring building (building 26, Fig. 1.7). However, the discovered architectural fittings enable certain conclusions. At least one door (or chest) was decorated with massive and ornate hinges; it is possible that this was the main door (tonovcov grad. Finds, Pl. 33: 1). Even though numerous keys and latches were discovered, no outer plates of locks were found. The movable wooden door and window wings were most likely attached with finer split hinges (tonovcov grad. Finds, Pl. 32: 3-24) that were not as durable, however the more massive ones were most likely harder to acquire or more demanding to produce. The wooden parts of the building and its fittings were held together by a large quantity of nails of varied length and a few clamps (tonovcov grad. Finds, Pls. 33: 4-9; 34-39). Only a few objects that could help us reconstruct the fittings in the building interior were discovered. As regards furniture we can be almost certain that they used chests of various sizes, in which they stored their possessions (see tonovcov grad. Finds, chapter 2.1.6; Pl. 33: 10-11). This form of furniture emerged in the Roman times and stuck almost until today. In numerous late Antique sites at least parts of the closing mechanism, if not also the metal fittings were discovered. In the cheaper versions the movable chest parts could be made from leather or wood. Remains of oil lamps were also discovered in the late Antiquity 2 layers (tonovcov grad. Finds, Pl. 65: 3-6), however as their shapes mainly indicate a 5th century origin (see tonovcov grad. Finds, chapter 4.1.2), we cannot state with any degree of certainty that they were connected to the life in the building, even though the possibility exists. Parts of hanging glass lamps were also discovered (see also tonovcov grad. Finds, chapter 3.1; Pl. 57: 19-21). ^66 617 667 618 619 668 669 . . • • • • 715 k. y / • '718 ^ V20 765 766 • ^^^ / ^^^ 767 768 A AsJ WeQ • 815 Ws • 819 820 866 ^^^ 868 869 • pasne spone in okovi / beit set parts • nakit / jeweilery • fibule / fibuiae 917 918 (I) 0 4m Sl. 3.15: Razprostranjenost pasnih spon in okovov, nakita in fibul v plasteh druge poznoantične faze v izkopnem polju stavbe 1. Fig. 3.15: Distribution of belt set parts, jewellery and fibulae in the Late Antiquity 2 layers in the excavation area of building 1. Po opustitvi hiše je s sten najprej odpadel omet (SE 08), potem pa so začeli propadati tudi zidovi. Gradivo, najdeno v plasteh druge poznoantične faze v stavbi 1 in njeni okolici kaže, da je bila ta zgrajena konec 5. oziroma v začetku 6. st. Tipično gradivo druge polovice 5. st. je izredno redko, tako da lahko domnevamo krajšo prekinitev poselitve med prvo in drugo poznoantično fazo. Na prekinitev kažejo tudi porušeni The majority of the everyday tasks probably took place in front of the building. The remaining pedestal for the pillar in front of wall 1 shows that the area was covered by a porch. A similar porch can be assumed on the other side of the entrance, alongside wall 5. Because the rock base is much higher at this location and partially levelled, the pillar could have been placed directly onto the rock. The intensive activities in the area in front of zidovi objektov prve poznoantične faze, katerih ruševina je bila ponovno uporabljena pri gradnji stavbe 2. 3.1.4 ŽGODNJI SREDNJI VEK potem ko je bila stavba 1 deloma že porušena, je bila še enkrat uporabljena za bivanje. Žgodnjesrednjeve-ške plasti so bile na izkopnem polju stavbe 1 ugotovljene samo v notranjosti objekta 1 in tu so bile zgoščene tudi the house were confirmed by the rich cultural layers next to the entrance, all of which could be dated to the second Late Antique phase (especially finds in quadrants 666, 716, 717 and 766, Fig. 3.15). In quadrant 666 the Late Antiquity 2 cultural layers included groups of flat slabs, the function of which has not been ascertained yet (Fig. 2.30). They were placed into very loose layers and they could be remains of some sort of paving in front of the building. As some of them were charred, we at first assumed that they represented a 618 819 668 669 Sl. 3.16: Razprostranjenost zgodnjesrednjeveških najdb v izkopnem polju stavbe 1. Fig. 3.16: Distribution of Early Medieval finds in the excavation area of building 1. Sl. 3.17: Najdbe iz zgodnjesrednjeveške plasti (SE 09). 1 novec; 2 bron; 3,4 železo; 5-17 keramika. M. 2-4 = 1:2; 5-17 = 1:3. Fig. 3.17: Finds from the Early Medieval layers (SU 09). 1 coin; 2 bronze; 3,4 iron; 5-17 pottery. Scale 2-4 = 1:2; 5-17 = 1:3. vse značilne zgodnjesrednjeveške najdbe (sl. 3.16). Gre za ruševinsko-žganinske plasti, ki so težko ločljive od plasti druge poznoantične faze in bolj kot na stalno poselitev kažejo na začasno zavetišče v delno porušeni poznoantični stavbi. Pasni okov iz poznoavarskega obdobja (Tonovcov grad. Najdbe, t. 8: 23) je bil najden v SE 11, ki predstavlja ruševinsko plast poznoantične stavbe. Večina ostalih predmetov, ki jih lahko opredelimo kot zgodnjesrednjeveške, je bila najdena v SE 10, ležeči nad SE 11. To so četverna modra jagoda (sl. 3.18: 10) in deli okovov nožnice noža (sl. 3.19: 1-3), ki omogočajo časovno opredelitev v konec 8. in začetek 9. stol. V tej plasti je bilo tudi precej grobe keramike (sl. 3.17: 6-11; 3.19: 22; 3.20: 1-13), ki časovno sicer ni ožje opredeljiva, vendar pa se od poznoantične razlikuje tako po obliki kot po fakturi (glej Tonovcov grad. Najdbe, pogl. 4.2.7). Poleg zgodnjesrednjeveških predmetov je bilo v plasteh zgodnjesrednjeveške faze najdenih tudi precej starejših elementov. Pozornost zbuja dejstvo, da med starejšimi najdbami iz teh plasti prevladujejo rimske najdbe. Med njimi so trije novci iz časa 1. in 2. st. (sl. 3.18: 6-8; glej tudi Tonovcov grad. Najdbe, pogl. 5.1, kat. št. 1, 2, 129) ter dva iz 4. st. (sl. 3.17: 1; 3.18: 9; glej tudi Tonovcov grad. Najdbe, pogl. 5.1, kat. št. 31, 151) in značilni predmeti prve poznoantične faze, kot sta obroč pasne spone z delfini (sl. 3.18: 14) in krožnik tipa Hayes 32/58 (sl. 3.19: 15). Predmeti druge poznoantične faze so redki (npr. odlomka vzhodnosredozemskih amfor, sl. 3.19: 16-17). Skelet (grob 1), ki je ležal ob zahodnem zidu prizidka (zid 10; sl. 2.33), je bil položen neposredno na skalno osnovo, ki ni bila vglobljena, in prekrit z ruševinskimi kamni. Ker ni grobne jame, verjetno lahko izključimo možnost, da je bil tam že v času poselitve hiše. Ker je bil prizidek zgrajen dokaj pozno (glej zgoraj) in je njegova ruševina prekrivala grob, lahko tudi nastanek slednjega domnevamo v času po koncu 6. st., natančneje pa ga ne moremo datirati. fireplace, however taking into account their density and the fact that most of them do not show any traces of heat this assumption was most likely erroneous. The rich late Antiquity 2 cultural layers were also found on the outer side of wall 4. At this the high numbers of Eastern Mediterranean amphorae that were found at the external corner of walls 3 and 4 (especially in quadrant 719) deserve a special mention. It is possible to assume that this was a storeroom for imported goods (see Tonovcov grad. Finds, chapter 4.1.8, Fig. 4.1). Amongst the pottery finds coarse kitchenware prevails (pots and bowls; see Tonovcov grad. Finds, chapter 4.2). The non-ceramic finds found inside and in the direct vicinity of building 1 are typical of self-sufficient autochthonus settlements in the Southeastern Alps: metal parts of furniture, tools, men and women's costume parts, weapons and stemmed goblets. Prestigious imported objects and objects of Germanic origin are rare (see Tonovcov grad. Finds, chapters 2.1 and 3.1). The most common coins found in the late Antiquity 2 layers are coins dating to the second half of the 4th and beginning of the 5th century (39.7 %; see Tonovcov grad. Finds, chapter 5.2). It is unlikely that the building was destroyed in a single disaster; more probably it fell into ruins slowly. The thick charred layer (SU 11) covered only the central part of the main room of building 1 (Figs. 2.9, 2.31), while the lack of charred remains elsewhere indicates that it is highly unlikely that the building was destroyed by a large fire. A gradual abandonment is also indicated by the fact that the house interior revealed very few objects from the time it was in use (Fig. 3.15). Once the building was abandoned the plaster slowly peeled of the walls (Su 08), and then the walls themselves started to decay. The material found in the late Antiquity 2 layers within building 1 and its surroundings shows that the building was constructed at the end of the 5th or beginning of the 6th century, and abandoned at the end of the 6th or at the beginning of the 7th century. Material that would be typical for the second half of the 5th century is extremely rare, thus we can assume a short break in the settlement between late Antiquity phases 1 and 2. This break is also indicated by the fact that the walls of the late Antiquity 1 building were demolished and their ruins were reused in the construction of building 2. 3.1.4 EARLY MEDIEVAL PERIOD Sl. 3.18: Najdbe iz zgodnjesrednjeveških plasti (SE 09 in 10). 1-5 keramika; 6-9 novci; 10,11 steklo; 14,16 bron; 12,13,15,1728 železo. M. 1-5 = 1:3; 10,11 = 1:1; 12-28 = 1:2. Fig. 3.18: Finds from the Early Medieval layers (SU 09 and 10). 1-5 pottery; 6-9 coins; 10,11 glass; 14,16 bronze; 1213,15,17-28 iron. Scale 1-5 = 1:3; 10,11 = 1:1; 12-28 = 1:2. once building 1 was already partially ruined, it was inhabited once again. The Early Medieval layers in the excavation area of building 1 were ascertained only inside building 1 and this was also where all typical Early Medieval finds were concentrated (Fig. 3.16). These are destruction and charred layers that are hard to Sl. 3.19: Najdbe iz zgodnjesrednjeveške plasti (SE 10). 1-14 železo; 15-22 keramika. M. 1-14 = 1:2; 15-22 = 1:3. Fig. 3.19: Finds from the Early Medieval layer (SU 10). 1-14 iron; 15-22 pottery. Scale 1-14 = 1:2; 15-22 = 1:3. Sl. 3.20: Najdbe iz zgodnjesrednjeveške plasti (SE 10). Vse keramika. M. = 1:3. Fig. 3.20: Finds from the Early Medieval layer (SU 10). All pottery. Scale = 1:3. SE/SU10 \ \ Sl. 3.21: Najdbe iz zgodnjesrednjeveške plasti (SE 10). Vse keramika. M. = 1:3. Fig. 3.21: Finds from the Early Medieval layer (SU 10). All pottery. Scale = 1:3. distinguish from the Late Antiquity 2 layers and do not indicate permanent settlement, but more a temporary refuge in the partially destroyed Late Antique building. The belt fitting from the Late Avar period (tonovcov grad. Finds, Pl. 8: 23) was found in SU 11, which represents the destruction layer of the Late Antique building. Most of the remaining objects that can be defined as Early Medieval were found in SU 10 which lay above SU 11. These are the four-partite blue bead (Fig. 3.18: 10) and parts of the fittings from a knife sheath (Fig. 3.19: 1-3) dated into the end of the 8th and beginning of the 9th century. This layer also revealed large quantities of coarse ware (Figs. 3.17: 6-11; 3.19: 22; 3.20: 1-13), which cannot be precisely dated, however it differs from Late Antique pottery by its shape as well as fabric (see tonovcov grad. Finds, chapter 4.2.7). Alongside the Early Medieval objects the layers belonging to the Early Medieval period revealed also numerous earlier finds. Attention was raised by the fact that Roman finds prevail amongst the early artefacts from these layers. These finds include coins dated to the 1st and 2nd century (Fig. 3.18: 6-8; see also tonovcov grad. Finds, chapter 5.1, Cat. Nos. 1, 2, 129) as well as two 4th century coins (Figs. 3.17: 1; 3.18: 9; see also tonovcov grad. Finds, chapter 5.1, Cat. Nos. 31, 151). Objects typical for Late Antiquity phase 1 were also common, such as for instance the belt buckle with dolphin heads (Fig. 3.18: 14) and a plate form Hayes 32/58 (Fig. 3.19: 15), while obj ects from Late Antiquity phase 2 were rare (for instance fragments of Eastern Mediterranean amphorae, Fig. 3.19: 16-17). The skeleton (grave 1) that was found alongside the western wall of the outhouse (wall 10; Fig. 2.33) was placed directly onto the bedrock, which was not deepened, and it was covered by destruction layer stones. As the skeleton was not placed into a grave pit we can probably exclude the possibility that it was there already at the time the building was in use. Because the outhouse was built rather late (see above), and the grave was covered by its ruins, we can assume that the grave was placed there after the end of the 6th century, however it is impossible to date it more precisely. 3.2 stavbi 2 in 3 3.2 buildings 2 and 3 v neposredni bližini sklopa cerkva so bili odkriti ostanki zidov, ki pripadajo dvema različnima stavbama. stavbi sta bili povezani med seboj, čeprav sta živeli v časovno različnih obdobjih. The discovered wall remains belong to two different buildings that were used in two different periods. First building 3 was erected, and from this a single wall was preserved in its entire length (wall 121), while walls su Sl. 3.22: Pogled z vzhoda na izkopno polje stavb 2 in 3. V ozadju sklop cerkva z zaščitnimi zgradbami nad severno in osrednjo cerkvijo. Fig. 3.22: Excavation area of buildings 2 and 3 - a view from the east. Ecclesiastical complex with protection buildings over the north and main church in the background. Najprej je bila postavljena stavba 3, od katere se je v celotni dolžini ohranil le en zid (zid 121), delno pa še nanj prizidana zidova (SE 126 in SE 127). Severni zid stavbe ni bil najden, vendar ni jasno, ali je bil popolnoma uničen, ali pa morda leži izven meja izkopnega polja. Na zid SE 121 so bili, ko je bil ta že delno porušen, prizidani zidovi stavbe 2 (sl. 2.34). Pod stavbama 2 in 3 so bili odkriti tudi naselbinski sledovi iz časa pred njuno izgradnjo. 126 and SU 127 that were attached to It were only partially preserved. The north wall was never discovered, however it is unclear whether it had been completely destroyed or whether it merely lies somewhere outside of the excavation area. Once wall SU 121 was already partially destroyed, the walls of building 2 were added to it (Fig. 2.34). Settlement traces from before buildings 2 and 3 were built can be found underneath the two buildings. 3.2.1 PRAZGODOVINA 3.2.1 PREHISTORIC PERIOD Sledovi prazgodovinske poselitve, odkriti v izkop-nem polju stavb 2 in 3, predstavljajo do sedaj najbolj bogate prazgodovinske ostanke na Tonovcovem gradu (tonovcov grad. Najdbe, pogl. 6.5). Tako so bili v plasti oranžne trde gline (SE 140 = 180) najdeni poškodovana glajena kamnita sekira, nož z naravnim hrbtom in obojestransko ploskovno retuširana puščična ost (Tonovcov grad. Najdbe, sl. 6.1: 1-2, 4). V isti plasti je bilo najdenih tudi nekaj zelo slabo ohranjenih, drobljivih, tipološko nedoločljivih fragmentov keramike. V SE 172, ki je poleg prazgodovinskih vsebovala tudi že antične najdbe, je bil najden neenakokrak trapez (Tonovcov grad. Najdbe, sl. 6.1: 3). Najdbe zajemajo široko časovno obdobje od srednjega paleolitika (nož z naravnim hrbtom) preko mezolitika (neenakokrak trapez) do eneolitika oziroma bronaste dobe (puščična ost, kamnita sekira; glej Tonovcov grad. Najdbe, pogl. 6.5). So far the traces of a prehistoric settlement discovered in the excavation area of buildings 2 and 3 represent the richest prehistoric settlement remains at tonovcov grad. The following prehistoric finds were discovered in the orange compact clay layer (SU 140 = 180): a damaged stone axe, a naturally backed knife and bifacially retouched arrow point (tonovcov grad. Finds, Fig. 6.1: 1-2, 4). The layer also included a few very poorly preserved and crumbly pottery fragments. An assimetrical trapezium (tonovcov grad. Finds, Fig. 6.1: 3) was also discovered in SU 172, which included Antique finds as well as prehistoric ones. the prehistoric finds appear through a broad spectrum of time, ranging from the middle Paleolithic period (naturally backed knife) over Mesolithic (assimetrical trapezium) to the Eneolithic or the Bronze Age (stone axe, arrow point; see tonovcov grad. Finds, chapter 6.5). 3.2.2 ANTIKA Nad prazgodovinskimi plastmi in pod ostanki stavbe 3, ki sodi v prvo poznoantično fazo, je bilo odkritih nekaj antičnih naselbinskih plasti, ki jih je težko natančneje časovno opredeliti, glede na stratigrafsko lego pa so vsekakor nastale pred začetkom prve po-znoantične faze. Nekaj kulturno skoraj sterilnih, verjetno erozijskih plasti (SE 176, 178) nad prazgodovinsko plastjo SE 180 kaže, da je prišlo do prekinitve življenja na tem območju. Prva plast, ki že vsebuje antične najdbe (ožje neopredeljiva keramika) je bila SE 179 v zahodnem delu izkopnega polja. Jama SE 177 je bila vkopana v SE 179 (sl. 2.37). V njej je bila najdena zgodnjerimska fibula Almgren 69, datirana v konec 1. in v 2. st. (Tonovcov grad. Najdbe, pogl. 2.2, t. 45: 1). To je do sedaj edini primer na Tonovcovem gradu, ko se zgo-dnjeantično gradivo tudi dejansko nahaja v plasti, ki pripada temu obdobju. Plast gline SE 173, ki je prekrivala jamo SE 177 in večino severnega in zahodnega dela izkopnega polja, je bila kulturno skoraj sterilna (v njej je bilo najdenih le 3.2.2 ANtlQUItY A few Antique settlement layers were discovered above the prehistoric layers and below the remains of building 3 (which is dated to late Antiquity phase 1). These layers are hard to date, however, taking their stratigraphic positions into account they certainly appeared before the beginning of late Antiquity 1 phase. Some culturally almost sterile, most likely erosion layers (SU 176, 178), above the prehistoric layer SU 180 indicate that there was a break in the settlement of this area. The first layer that already includes Antique finds (pottery that cannot be more precisely defined) was found in SU 179 in the western part of the excavation area. Pit SU 177 was dug into SU 179 (Fig. 2.37). The pit revealed an Early Roman fibula Almgren 69, that could be dated into the end of the 1st and the 2nd century (see tonovcov grad. Finds, chapter 2.2, Pl. 45: 1). This is so far the only example at tonovcov grad, in which Early Antique material is actually found in a layer that b elongs to this period. The clay layer SU 173, which covered pit SU 177 and most of the north and west part of the excavation nekaj živalskih kosti). Plast verjetno predstavlja določeno prekinitev življenja na tem območju. v SE 173 je bila vkopana jama SE 175 (sl. 2.38). Segala je skozi SE 173 še v spodaj ležečo SE 179 (sl. 2.37), vsebovala pa je keramiko in živalske kosti. Od opredeljivih keramičnih oblik je bil v njej najden lonec tipa 3, na Tonovcovem gradu značilen za prvo poznoantično fazo (Tonovcov grad. Najdbe, t. 102: 6). v skrajnem severnem delu, že na meji izkopnega polja, je na SE 173 stala skupina večjih kamnov (SE 171; sl. 2.38,2.39). Ali predstavljajo ti kamni ostanek zidu, ne moremo ne potrditi ne zanikati, saj je bil izkop na tem mestu zaključen. Če gre za zid, je ta starejši od zidov stavbe 3, saj ga prekriva SE 160 (sl. 2.39), ki pa poteka pod zidovoma SE 121 in 126 (sl. 2.35). v jugozahodnem vogalu stavbe 3 je nad SE 173 ležala do 30 cm debela, precej kulturna plast SE 172 (sl. 2.38), ki pa je vsebovala časovno zelo mešane najdbe. Tako so bili skupaj najdeni novec cesarja Aurelijana (leto 272, glej Tonovcov grad. Najdbe, pogl. 5, kat. št. 22) in kamnito orodje (Tonovcov grad. Najdbe, sl. 6.1: 3). Plasti, ki so ležale pod ostanki stavbe 3, lahko glede na najdbe postavimo v čas od 1. do 3. st. Ker pa so bili predmeti iz tega časa na Tonovcovem gradu pogosto najdeni tudi v mlajših plasteh, je predvsem stratigrafska lega pod plastmi prve poznoantične faze tista, ki dovoljuje opredelitev starejših plasti v obdobje pred drugo polovico 4. st. 3.2.3 PRVA POZNOANTIČNA FAZA (PA 1) v tej fazi je bila postavljena stavba 3. Njeni ohranjeni zidovi (SE 121, 126 in 127) so bili postavljeni na SE 172, torej na plast, ki vsebuje mešano prazgodovinsko in antično gradivo. Skoraj vso notranjost stavbe 3 znotraj ohranjenih zidov je zavzemala SE 160 (sl. 2.40). Plast ni bila zelo debela, najdbe iz nje so skromne in kronološko slabo opredeljive (železen predmet, nekaj atipičnih odlomkov keramike). Plast sodi v čas uporabe stavbe 3. Vse ostale plasti, ki bi še lahko pripadale temu obdobju, so bile očitno uničene pri gradnji stavbe 2. Močna kulturna plast (SE 150 = 144; sl. 2.41), ki prekriva ostanke vzhodnega (SE 127) in zahodnega (SE 126) zidu stavbe 3 in njen osrednji del nad SE 160, namreč vsebuje gradivo, ki ga lahko datiramo v 6. st. in začetek 7. st. in je povezano z življenjem stavbe 2. Kljub skromnim najdbam lahko glede na stra-tigrafsko lego nad antičnimi plastmi in pod plastmi faze PA 2 domnevamo, da je stavba 3 živela v času prve poznoantične faze, to je v drugi polovici 4. in v začetku 5. st. Opuščena je bila verjetno do sredine 5. st. Do konca 5. st. je bila očitno že skoraj popolnoma porušena, kajti takrat so njeno ruševino in edini delno še stoječ zid (SE 121) uporabili pri gradnji nove stavbe 2. area, was almost culturally sterile (it included only a few animal bones). This layer is likely to present a certain break in activities in this area. Pit SU 175 was dug into SU 173 (Fig. 2.38). It reached through SU 173 into the underlying SU 179 (Fig. 2.37) and it included pottery and animal bones. From the definable pottery forms a pot type 3 was found, and this is at tonovcov grad typical for the Late Antiquity phase 1 (tonovcov grad. Finds, Pl. 102: 6). In the far north of the excavation area, already on its border, SU 173 was covered by a group of large stones (SU 171; Figs. 2.38,2.39). As the excavation ended at this point it is impossible to confirm or deny whether these stones represent wall remains. If these are wall remains, the wall had to be older than the walls of building 3 as it was covered by SU 160 (Fig. 2.39), which run under walls SU 121 and 126 (Fig. 2.35). In the southwest corner of building 3 layer SU 173 was covered by a 30 cm thick, culturally rich layer SU 172 (Fig. 2.38), however this layer included a diverse mixture of finds from various periods. Thus a coin of the Emperor Aurelius (272 AD, see tonovcov grad. Finds, chapter 5, Cat. No. 22. ) and a stone tool were all found together (tonovcov grad. Finds, Fig. 6.1: 3). The finds discovered in layers below the remains of building 3 can be placed in the period between the 1st and 3rd century. However, because the objects from this period were often found at tonovcov grad also in later layers, the stratigraphic position under the Late Antiquity 1 layers allows for these earlier ones to be placed prior to the second half of the 4th century. 3.2.3 LATE ANTIQUITY PHASE 1 (LA 1) Building 3 was erected during this phase. Its preserved walls (SU 121, 126 and 127) were built on SU 172, i.e. on a layer that includes a mixture of prehistoric and Antique material. Within the preserved walls almost the entire interior of building 3 was covered by SU 160 (Fig. 2.40). The layer was not very thick, and the finds from it were very modest and chronologically hard to define (a metal object, some atypical ceramic fragments). The layer belongs into the period in which building 3 was used. All other layers that could belong to this period were obviously ruined during the construction of building 2. The strong cultural layer (SU 150 = 144; Fig. 2.41) that covers the remains of the east (SU 127) and west (SU 126) wall of building 3 and its central part above SU 160 includes material that can be dated into the 6th and the beginning of the 7th century and is linked to the time in which building 2 was in use. Regardless of the modest finds the stratigraphic position above the Roman layers and under the LA 2 layers leads us to believe that the building 3 was used 3.2.4 DRUGA POZNOANTIČNA FAZA Konec 5. ali začetek 6. st. verjetno pomeni čas izgradnje stavbe 2. Drobne najdbe iz notranjosti stavbe 2 so izredno skromne in kronološko slabo izpovedne, izvirajo pa pretežno iz ruševinskih plasti. Veliko bogatejše in bolj izpovedne so plasti severno od stavbe 2, ki so prekrivale ostanke stavbe 3, spadajo pa v čas življenja stavbe 2. Način gradnje stavbe 2 se precej razlikuje od gradnje ostalih raziskanih zgradb na Tonovcovem gradu. Za severni zid nove stavbe 2 so izkoristili očitno najbolje ohranjeni, južni zid starejše stavbe 3 (SE 121), na katerega so nato v tehniki suhozida prizidali zidova SE 122 in SE 124. Tudi južni zid stavbe (SE 123) je bil zidan brez uporabe malte. Za zidavo so verjetno uporabili ruševino starejše stavbe 3. Temelji prizidanih zidov so postavljeni približno 0,5 m višje kot temelj osnovnega zidu SE 121. Podoben način gradnje je bil doslej poznan le pri stranskem prostoru (prizidku) stavbe 1, vendar je bil stranski prostor stavbi 1 prizidan še v času njenega življenja (glej pogl. 3.1.3). Najdbe iz notranjosti stavbe so zelo skromne. Na vrhu prazgodovinske plasti SE 140 = SE 180 je ležal velik del ostenja tipološko ožje nedoločljive zgodnjeantične amfore (glej Tonovcov grad. Najdbe, t. 100: 8), poleg pa precej ostankov tegul, ki so na območje stavbe 2 prišle iz nekaj višje ležeče severne cerkve. Tako gre tudi pri tej amfori verjetno za residualno najdbo. Možno je celo, da je bila amfora že prej sekundarno uporabljena kot strešnik. Najbolj številne so bile najdbe v SE 106 in SE 112. V 6. st jih opredeljuje nekaj značilnih predmetov, npr. ročaj steklene svetilke (Tonovcov grad. Najdbe, t. 58: 18), tri noge steklenih kozarcev (Tonovcov grad. Najdbe, t. 58: 12-14) in nekaj vzhodnomediteranskih amfor (Tonovcov grad. Najdbe, t. 100: 6-7). V obeh plasteh je bilo najdeno tudi precej strešnikov. Ker sodijo steklene svetilke in kozarčki med značilen inventar cerkva, lahko domnevamo, da so na območje stavbe 2 prišli, podobno kot strešniki, iz območja severne cerkve. Od grobe keramike je bilo v stavbi 2 najdenih le nekaj tipološko neopredeljivih kosov. Bolj bogate kot v notranjosti pa so bile najdbe v plasteh, ki pripadajo drugi poznoantični fazi severno od zidu SE 121. Tu so ostanke stavbe 3 prekrivale kulturne plasti iz časa življenja stavbe 2. V osrednjem delu izkopnega polja sta bili v SE 160 vkopani dve jami nepravilne oblike (SE 161 in 162, sl. 2.40). V SE 161 so bili najdeni fragmenti roženega glavnika, nož in železen predmet, ustje steklenega kozarčka, fragment lonca tipa 4a, značilen za fazo PA 1 in fragment egejske amfore (sl. 3.23: 1-8). V SE 162 pa so bili najdeni dvoramna fibula, značilna za drugo polovico 6. in prvo polovico 7. st (glej Tonovcov grad. during Late Antiquity 1 phase, i.e. in the second half of the 4th and beginning of the 5 th century. It was probably abandoned by the mid 5th century. by the end of the 5th century it was most likely already completely destroyed, for this was the time when the only remaining wall (SU 121) that was found under its ruins was used in the construction of the new building 2. 3.2.4 LATE ANTIQUITY PHASE 2 Building 2 was most likely built at the end of the 5th or beginning of the 6th century. The small finds from the interior of building 2 are extremely modest and chronologically poorly explained, and most of them were found in the destruction layers. However, there are much richer and clearer layers outside of building 2. They covered the remains of building 3 and belong to the period in which building 2 was in use. Building 2 was built in a different way compared to the remaining researched buildings at Tonovcov grad. For the north wall of the new building 2 they used the best preserved, south wall of the older building 3 (SU 121), onto which they added walls SU 122 and SU 124 in the dry masonry technique. The south wall (SU 123) was also built without the use of mortar. It is likely that they used the ruins of the older building 3 for the construction. The foundations of the added walls were placed approximately 0.5 m higher than the foundations of the primary wall SU 121. A similar type of construction has so far been noticed only at the outhouse of building 1, however the outhouse was added to building 1 while this was still in use (see chapter 3.1.3). The finds from the building itself were very modest. On the top of the prehistoric layer SU 140 = SU 180 lay a large fragment of Early Antique amphora that cannot be more precisely defined typologically (see Tonovcov grad. Finds, Pl. 100: 8). Next to it quite a few tegulae remains were found, and for them it is believed that they came to the area of building 2 from the somewhat higher lying north church. Thus this amphora is most likely a residual find. It is even possible that the amphora was already in its secondary use as a roof tile. SU 106 and SU 112 were the richest in finds. Some characteristic objects place it into the 6th century, such as for instance the handle of the glass lamp (Tonovcov grad. Finds, Pl. 58: 18), three glass stems (Tonovcov grad. Finds, Pl. 58: 12-14) and some Eastern Mediterranean amphorae (Tonovcov grad. Finds, Pl. 106: 6-7). Both layers also revealed quite a large number of roof tiles. As glass lamps and vessels are typical church inventory, we can assume that they came into the area of building 2 from the area of the north church (similarly as the roof tiles). Building 2 revealed only a few typologically indefinable coarse ware pieces. Sl. 3.23: Stavba 3, najdbe iz SE 161 in 162. 1,2 železo; 3 rogovina in železo; 4,10 steklo; 5-8,11 keramika; 9 bron. M. 1-4,9,10 = 1:2; 5-8,11 = 1:3. Fig. 3.23: Building 3, finds from SU 161 and 162. 1,2 iron; 3 antler and iron; 4,10 glass; 5-8,11 pottery; 9 bronze. Scale 1-4,9,10 = 1:2; 5-8,11 = 1:3. Najdbe, pogl. 2.2), ustje steklenega kozarčka in fragment egejske amfore (sl. 3.23: 9-11). Jami sta verjetno nastali hkrati (v obeh sta bila najdena dva odlomka iste amfore), vsebujeta pa časovno mešano gradivo. Severno od stavbe 2 so na vrhu SE 160 in obeh vkopov (SE 161 in 162) ležale kamnite plošče. Bile so postavljene posamično ali v skupinah (sl. 2.40). Osrednji del izkopnega polja nad SE 160 je prekrivala plast črne, močno kulturne prsti (SE 150 = 144; sl. 2.41). Tudi v SE 150 smo lahko še vedno sledili skupinam kamnitih plošč, prvič dokumentiranim v SE 160. Podobne kamnite plošče so bile najdeni tudi na zunanji, vhodni strani stavbe 1 (glej pogl. 2.3 in 3.1). Tako tam kot pri stavbi 2 njihova funkcija ni jasna. The finds in the late Antiquity 2 layers, north of wall 121, were much richer that the finds from the interior. Here the remains of building 3 covered the cultural layers from the time building 2 was in use. In the central part of the excavation area two irregular shaped pits (SU 161 and 162, Fig. 2.40) were dug into SU 160. SU 161 revealed fragments of an antler comb, a knife and a metal object, a glass rim, a fragment of a type 4a pot (typical for phase lA 1) and a fragment of an Aegean amphora (Fig. 3.23: 1-8). SU 162 revealed an equal-arm fibula typical of the second half of the 6th and the first half of the 7th century (see tonovcov grad. Finds, chapter 2.2), a glass rim and a fragment of an Aegean amphora (Fig. 3.23: 9-11). It is likely that both Kljub slabi časovni opredeljivosti večine najdb lahko SE 150 datiramo v 6. st. V ta čas sodi nekaj fragmentov steklenega posodje in keramike, od kovinskih predmetov pa npr. ročaj ščita (Tonovcov grad, Najdbe, t. 46: 9). Vprašanje funkcije stavbe 2 glede na skromne ostanke v njeni notranjosti ostaja odprto. Ker v stavbi ni bilo najdenih ostankov ognjišča, bi težko domnevali, da gre za bivalni objekt. Tudi skoraj popolna odsotnost drobnih najdb naselbinskega značaja v notranjosti govori za nebivalni značaj stavbe. Po drugi strani so bile dokaj bogate drobne najdbe iz časa 6. in začetka 7. st. na njeni severni zunanji strani, na območju starejše stavbe 3 (SE 150). Nekaj kosov orodja (dve punci/zlatarsko orodje, nakovalce: Tonovcov grad. Najdbe, t. 43: 6,11; 45: 34), več koščkov bronaste in svinčene pločevine, stari predmeti (krilat pasni okov, Tonovcov grad. Najdbe, t. 45: 15) in polizdelek sponke verižice (Tonovcov grad. Najdbe, t. 45: 3) bi morda lahko kazali na izdelavo ali vsaj predelavo manjših bronastih predmetov na tem prostoru (Tonovcov grad. Najdbe, pogl. 2.2). Med temi najdbami pretežno naselbinskega značaja izstopa ročaj ščita (Tonovcov grad. Najdbe, t. 46: 9). Najdbe iz ruševinsko-humusnih plasti znotraj stavbe 2 so skromne in časovno slabo opredeljive. Spet izstopata najdbi orožja - železne bradate sekire in puščične osti (Tonovcov grad. Najdbe, t. 43: 1; 45: 18; glej tudi pogl. 2.2). Podobno razporeditev najdb, se pravi večinoma izven same stavbe, opažamo tudi pri stavbi 1, le da tam pomanjkanje najdb v notranjosti ni tako očitno kot pri stavbi 2. Tudi pri stavbi 1 pa je bila večina najdb skon-centriranih v plasteh zunaj samega objekta, predvsem pred vhodom. Tam so plasti 6. st. ravno tako prekrivale ostanke starejšega zidu (glej pogl. 2.3 in 3.1). Zaradi lokacije stavbe 2 v bližini cerkva je bila že pred pričetkom raziskav postavljena teza, da gre pri tej zgradbi za baptisterij (glej pogl. 4.2.6). Tezo bi na podlagi najdb težko potrdili, proti njej govori tudi razmeroma slaba kvaliteta gradnje stavbe 2. Kljub temu jo ob dejstvu, da v samem cerkvenem sklopu baptisterij ni bil zanesljivo potrjen, tudi ne moremo povsem zanikati (glej pogl. 4.2.6). 3.2.5 ZGODNJI SREDNJI VEK V to fazo zanesljivo lahko postavimo samo grob št. 18 (sl. 2.43). V njem sta bila najdena jermenski zaključek (Tonovcov grad. Najdbe, t. 44: 24), okvirno datiran v 8. st. in svitkasta jagoda (Tonovcov grad. Najdbe, t. 44: 23; sl. 2.10: 2), datirana v konec 7. in začetek 8. st. (glej Tonovcov grad. Najdbe, pogl. 2.2). Ostali trije grobovi za stavbo 2 so bili sicer brez pridatkov, vendar zaradi njihove dokaj enotne usmeritve domnevamo, da jih lahko datiramo podobno kot grob 18. Vsekakor so bili za južni zid stavbe 2 zakopani v pits were dug at the same time as they both included fragments of the same amphora as well as materials that came from different periods. stone slabs were found north of building 2, on the top of su 160 and on both pits (SU 161 and 162). They were positioned individually or in groups (Fig. 2.40). The central part of the excavation area above SU 160 was covered by a layer of black, culturally rich soil (SU 150 = 144; Fig. 2.41). SU 150 also included groups of stone slabs that were first documented in SU 160. Similar stone slabs were discovered also on the exterior, at the entrance to building 1 (see chapters 2.3 and 3.1). Once again their function is unclear (similar as in building 2). Regardless of the fact that it is hard to date most finds SU 150 can be dated into the 6th century. A few glass and pottery fragments belong into this period, as does the shield handle (tonovcov grad. Finds, Pl. 46: 9). taking into account the modest remains in the interior the function of building 2 remains unexplained. The lack of any fireplace remains within the building makes it hard to believe that it was used for living quarters. This idea is only strengthened by the almost non-existent small finds that are typically found in settlements. on the other hand relatively rich small finds from the 6th and beginning of the 7th century were discovered alongside its northern exterior wall, in the area of the older building 3 (SU 150). Some tools (a goldsmith's tool, anvil: tonovcov grad. Finds, Pls. 43: 6,11; 45: 24), numerous pieces of bronze and lead, old objects (belt fitting, tonovcov grad. Finds, Pl. 45: 15) and a semi product of a necklace clasp (tonovcov grad. Finds, Pl. 45: 3) could indicate that small bronze objects were manufactured or at least reworked here (see tonovcov grad. Finds, chapter 2.2). Amongst these finds, most of which have a typically settlement character, the shield handle stands out (tonovcov grad. Finds, Pl. 46: 9). The finds from the destruction layers of building 2 are rare and hard to date. once again the weapon finds stand out - an axe and an arrowhead (tonovcov grad. Finds, Pls. 43: 1; 45: 18; see also chapter 2.2). Similar distribution of finds was also uncovered in building 1, only that there the lack of finds in the interior was not as obvious as in building 2. In the excavation area of building 1 most finds were concentrated in the layer outside of the building itself, mainly in front of the entrance. The remains of the older wall were covered by 6th century layers (see chapters 2.3 and 3.1). As building 2 was so close to the churches it was assumed already before the excavations started that this building was a baptistery (see chapter 4.2.6). It is hard to back this hypothesis with the finds, and the relatively poor construction of building 2 also speaks against this idea. However, as a baptistery has not been reliably confirmed in the ecclesiastical complex, this hypothesis cannot be entirely neglected (see chapter 4.2.6). času, ko je stavba vsaj v osnovi še stala. vkopani so bili v prazgodovinsko plast sE 140 oziroma v geološko osnovo se 117, ruševina stavbe jih je popolnoma prekrila, v njej pa ni bilo opaziti nobenega vkopa, ki bi govoril za to, da bi jih pokopali potem, ko je bila stavba že porušena. Tudi pri grobovih se kaže podobnost s stavbo 1, kjer je bil za zid prizidka na skalno osnovo položen skelet, ki so ga potem prekrili kamni porušenega zidu (glej pogl. 2.3.4). 3.2.5 EARLY MEDIEVAL PERIOD only grave No. 18 (Fig. 2.43) can be reliably dated into this phase. The grave revealed a strap end (tonovcov grad. Finds, Pl. 44: 24) that could be roughly dated into the 8th century and a green bead (tonovcov grad. Finds, Pl. 44: 23; Fig. 2.10: 2) that could be dated to the end of the 7th or beginning of the 8th century (see tonovcov grad. Finds, chapter 2.2). No grave goods were discovered in the remaining three graves behind building 2, however due to their relatively unified orientation we can assume that they can be roughly dated into the same period as grave 18. It is certain that they were dug behind the south wall in a period in which the building was still standing. They were dug into the prehistoric layer su 140 or into the geological base su 117, and completely covered by the building ruins, which did not reveal any traces of excavation that would indicate that these individuals were buried once the building was already demolished. The graves also show similarities with building 1, where a skeleton was placed on the bedrock behind the wall of the outhouse, and then covered by the rocks of the destroyed wall (see chapter 2.3.4). 3.3 sklop cerkva 3.3 the ecclesiastical complex Vse tri cerkve na Tonovcovem gradu sodijo med preproste pravokotne enoladijske cerkve s prosto stoječo klopjo za duhovnike. V celoti so bile grajene iz kamnitih lomljencev in prekrite z opeko. Severna in osrednja cerkev sta bili med seboj povezani v celotni dolžini, prezbiterij južne cerkve pa je s prezbiterijem osrednje povezoval štirikoten zidan vmesni prostor. Prostor med ladjama osrednje in južne cerkve je bil vglobljen v skalno osnovo, vendar nepozidan. Orientacija južne cerkve nekoliko odstopa od orientacije severne in osrednje. Vse tri cerkve so imele prizidane nartekse (sl. 2.47; pril. 4). Cerkveni sklop kaže več faz dozidav in prezidav. 3.3.1 SEVERNA IN OSREDNJA CERKEV Najprej sta bili postavljeni severna in osrednja cerkev. Kot kaže velik kamnit blok, ki povezuje vzhodni stranici obeh cerkva, sta bili zgrajeni hkrati (sl. 2.51, pogl. 2.5.1, 2.5.2). Cerkvi sta bili med seboj povezani z dvema prehodoma - med ladjama in v prezbiterijskem delu. Prezbiterija sta bila od ladje ločena z zidano pregrado. Notranje stene so bile ometane. Kako je bila videti notranjost v začetni gradbeni fazi, ne moremo v celoti rekonstruirati, ker je bila uničena s kasnejšimi prezidavami, pa tudi raziskana površina plasti pod sedaj ohranjenimi estrihi je bila majhna (glej pogl. 2.5). S precejšnjo gotovostjo pa lahko rečemo, da v začetni gradbeni fazi cerkvi nista imeli estrihov, glede na ostanke strešnikov v plasteh pod danes ohranjenimi estrihi (SE 28, 32, 58, 59) pa lahko sklepamo, da sta bili že na začetku kriti z opeko. Duhovniške klopi iz prve gradbene faze niso ohranjene, ravno tako ne oltarji in stopnice med ladjama in prezbiterijema. Časa nastanka cerkvenih zgradb natančno ne moremo določiti, saj so najdbe v plasteh pod estrihi skromne in kronološko slabo izpovedne (sl. 3.24). V severni cerkvi je bilo v SE 58 najdeno dno afriškega sigilatnega krožnika (sl. 3.24: 3), datirano v drugo polovico 4. in v začetek 5. st. (glej Tonovcov grad. Najdbe, pogl. 4.1.1), kar pa seveda ne pomeni tudi datacije plasti. Drugi All three churches at Tonovcov grad are simple rectangular single nave churches with a free standing clergy bank. They were constructed from quarry stones and covered by tegulae. The north and main churches were joined in their entire length, while the presbytery of the south one was linked with the presbytery of the main church by a square stone structure. The area between the naves of the main and south churches was carved into the bedrock, but did not contain any built structures. The orientation of the south church differs slightly from the orientations of the north and main churches. All three churches had narthices (Fig. 2.47; Insert 4). Parts of the ecclesiastical complex were added or remodelled in various phases. 3.3.1 THE NORTH AND MAIN CHURCHES The north and the main church were the first to be built. The large stone block that links the east sides of the two churches indicates that they were built at the same time (Fig. 2.51, chapters 2.5.1, 2.5.2). The two churches were joined by two passages - one linked the two naves and the other linked the two presbyteries. A partition wall separated the presbyteries from the naves. All interior walls were covered in plaster. It is impossible to reconstruct the interior during its first construction phase, for it was destroyed by the later additions, and the researched area of layers under the preserved mortar floors is small (see chapter 2.5). However, we can state with relative certainty that there were no mortar floors during the first construction phase. The tegulae discovered in the layers under the preserved mortar floors (SU 28, 32, 58, 59) reveal that churches were covered by tegulae from the very beginning. The clergy banks from the first construction phase are not preserved, nor are the altars and stairs between the naves and presbyteries. It is impossible to precisely define when churches were built, for the finds in the layers under the mortar floors are modest and difficult to date (Fig. 3.24). In the north church, in SU 58, the bottom part of an African Red Sl. 3.24: Sklop cerkva, najdbe iz plasti pod hodnimi površinami. A - severna cerkev, B - osrednja cerkev, C - narteks; 1-2, 6 železo; 3-5, 10 keramika; 7 bron; 8-9,11 steklo. M. 1-2, 6-11 = 1:2; 3-5 = 1:3. Fig. 3.24: Ecclesiastical complex, finds from layers underneath floor levels. A - north church, B - main church, C - narthex; 1-2, 6 iron; 3-5, 10 pottery; 7 bronze; 8-9, 11 glass. Scale 1-2, 6-11 = 1:2; 3-5 = 1:3. časovno izpoveden kos je bil najden v narteksu severne cerkve, pod grobom 13, v SE 56. Gre za nogo steklenega kozarca (sl. 3.24: 11), kakršni so bili na jugovzhodnoalpskem območju v uporabi od poznega 5. st. dalje (glej tudi Tonovcov grad. Najdbe, pogl. 3.3, 3.7.2). v drugi fazi so bili v obeh cerkvah narejeni estrihi, postavljeni duhovniški klopi in oltarja ter sezidane stopnice iz ladij v prezbiterija, v severni cerkvi pa postavljena še prižnica (ambon). Kot kažejo nekateri ohranjeni detajli, so bili vsi ti elementi dodani hkrati. v severni cerkvi je vidno, da estrih iz tal neposredno prehaja v omet klopi oziroma oltarja (glej pogl. 2.5.1, sl. 2.54). Tudi stopnice v obeh ladjah so bile prizidane na že ometano cerkveno steno (sl. 2.66). Ostanki estriha, s katerim je bila premazana, so bili ohranjeni tudi na vrhu prižnice. Slip Ware plate (Fig. 3.24: 3) dated into the second half of the 4th or beginning of the 5th century (see Tonovcov grad. Finds, chapter 4.1.1) was discovered. Of course, this does not necessarily represent the date of the layer. The second piece that provided information as regards the date of construction was discovered in the narthex of the north church, under grave 13, in SU 56. This was a stemmed goblet foot (Fig. 3.24: 11), which was in use in the southeastern Alpine area from the late 5th century onwards (see also Tonovcov grad. Finds, chapters 3.3, 3.7.2). The mortar floors were laid in both churches during the second construction phase. During this phase the clergy banks and altars and the stairs from the naves to the presbyteries were also built as was the pulpit in the north church. Certain preserved details reveal that all of these elements were added at the same time. In the Kamnita pregrada, ki je ločevala prezbiterija od ladij, je bila postavljena na naravni skalni stopnici. V severni cerkvi je bila ohranjena višina pregrade nad hodnim nivojem ladje največ 60 cm, nad hodnim nivojem prezbiterija pa 10-20 cm. Na nekaterih mestih je bil še ohranjen estrih, premazan čez vrh pregrade, tako da lahko sklepamo, da je ohranjena višina tudi dejanska višina zidu pregrade. Odkritih ni bilo nobenih ostankov vdolbin oziroma lukenj, pač pa je bilo v osrednji cerkvi v prezbiteriju in ladji najdeno večje število velikih železnih žebljev, ki bi lahko pripadali leseni nadgradnji (Tonovcov grad. Najdbe, sl. 3.2). Ohranjena pregrada v osrednji cerkvi je sicer nižja kot tista v severni in sega okrog 30 cm nad hodni nivo ladje in le nekaj cm nad hodni nivo prezbiterija, vendar ni jasno, ali je bila to tudi njena dejanska višina ali pa je morda segala višje. Obe duhovniški klopi sta bili prosto stoječi, sestavljeni iz sedala, naslonjala in podstavka za noge, v sredini pa sta imeli katedro, do katere so vodile stopnice. Klop v severni cerkvi je bila prizidana na že prej ometan zid (glej pogl. 2.5.1). Obe klopi sta bili ometani najprej z grobim, nato pa prevlečeni še s finim ometom, iz katerega so gledali le vrhovi kamnov (sl. 2.57). Stranici klopi v severni cerkvi sta segali povsem do cerkvenih zidov (zidova 1 in 3), tako da je za njima nastal zaprt prostor, stranici klopi v osrednji cerkvi pa sta potekali vzporedno z zidovoma cerkve (zidova 3 in 6), tako da je med njimi ostal prazen prostor. Za stranicama klopi osrednje cerkve je bilo najdenih precej ostankov steklenih posodic in svetilk (Tonovcov grad. Najdbe, sl. 3.2; t. 59: 15-16,18,22; 60: 2,5-6,8,10,12). Morda so prostor uporabljali za shrambo predmetov, potrebnih ob liturgiji, lahko pa lučke predstavljajo tudi ostanek osvetlitve ob duhovniški klopi. Obe cerkvi sta imeli v osrednjem delu prezbiterija zidana oltarja, v severni cerkvi pa so bili ob južnem delu pregrade med ladjo in prezbiterijem ohranjeni tudi ostanki zidane prižnice (ambona). Ob oltarju severne cerkve je bil v ruševini najden del rimske kamnite žare, ki je bila lahko uporabljena kot relikviarij (sl. 2.58). Prezbiterij severne cerkve je bil prekrit z estrihom, v osrednji cerkvi pa ostanki estriha niso bili najdeni. Hodno površino je tu predstavljala umetno zravnana skalna osnova oziroma plast rumeno-rjave ilovnate zemlje (SE 26), ki je prekrivala neravnine v skali in zapolnjevala skalne razpoke. Za določitev časa prenove cerkva je pomembna najdba male amforice (spatejona) v prezbiteriju osrednje cerkve. v eno od skalnih razpok, ki je bila še dodatno razširjena, segala pa je pod oltar, je bila položena mala amfora, v kateri je bil Justinijanov srebrnik (sl. 2.65; Tonovcov grad. Najdbe, t. 101: 7). Tako čas izdelave srebrnika (540-552, glej Tonovcov grad. Najdbe, pogl. 5.1, kat. št. 157) kot časovna umestitev spatejona (Tonovcov grad. Najdbe, pogl. 4.1.7) datirata to zadnjo preureditev cerkva, ko so bili narejeni estrihi ter postavljeni klopi in oltarji, v sredino 6. st. north church the mortar continues from the floor into the plaster that covers the bank and the altar (see chapter 2.5.1, Fig. 2.54). The stairs in both naves were added to the already plastered church walls (Fig. 2.66). The remains of the mortar were preserved also at the top of the pulpit. The stone partition wall that separated the presbyteries from the naves was placed on a natural rock edge. In the north church the height of the preserved partition wall reached up to a maximum of 60 cm above the floor level of the nave and between 10 and 20 cm above the floor level of the presbytery. In certain spots the mortar that covered the top of the partition wall was preserved which indicates that the preserved height was also the actual height of the partition wall. No holes or dents were discovered, however, a number of long nails that could have belonged to a wooden superstructure were discovered in the main church's presbytery and nave (Tonovcov grad. Finds, Fig. 3.2). The preserved partition wall in the main church is lower than the one in the north church and reaches approximately 30 cm above the floor level in the nave and only a few centimetres above the floor level in the presbytery. However, it is unclear whether this was its actual height, or did it reach somewhat higher. Both clergy banks were free standing and they consisted of a seat, a backrest and a footstool, while in the centre steps led to the cathedra. The bank in the north church was attached to a previously plastered wall (see chapter 2.5.1). Both banks were covered in a layer of coarse plaster, which was in turn covered by a layer of fine plaster, from which only the stone tops peaked through (Fig. 2.57). The sides of the bank in the north church extended all the way to the church walls (walls 1 and 3), thus creating an enclosed space behind them, while the sides of the bank in the central church ran parallel to the church walls (walls 3 and 6), thus creating an empty space between the two. Numerous glass vessel and lamp remains were found behind the bank in the main church (Tonovcov grad. Finds, Fig. 3.2; Pls. 59: 15-16,18,22; 60: 2,5-6,8,10,12). It is possible that the space was used for storing liturgy objects, however the lamps could also be the remains of the lighting used for the clergy bank. Both churches revealed masonry altars in the centre of their presbyteries, and the north church also revealed remains of a masonry pulpit (ambo) that was discovered alongside the south partition wall between the nave and the presbytery. The ruins alongside the altar in the north church revealed a fragment of a stone Roman urn which could have been used as a reliquary (Fig. 2.58). While the presbytery of the north church was covered by a mortar floor, the main church had no similar floor. Here the floor level was represented by the artificially levelled bedrock or a layer of yellowish-brown soil (SU 26) that covered the irregularities and filled the gaps in the rock. The small amphora (spatheion) found in the presbytery of the main church offered an important clue as Skalna razpoka naravnega izvora, zapolnjena z antičnim gradivom, ki poteka po vzdolžni osi ladje osrednje cerkve in se pod pregrado prezbiterija nekoliko razširi, bi lahko bila uporabljana kot piscina baptisterija (glej pogl. 4.2.6). Pregrada na tem mestu ni ohranjena, lahko, da je tudi ni bilo (pril. 2). Jama groba 5 v severozahodnem vogalu osrednje cerkve je prebila estrih v cerkveni ladji (pril. 2), torej lahko ta pokop postavimo v čas po ureditvi estrihov, po sredini 6. st. Glede na količino najdenih odlomkov okenskega stekla lahko rečemo, da so bila okna cerkva v zadnji gradbeni fazi zastekljena (Tonovcov grad. Najdbe, sl. 3.3). Razprostranjenost odlomkov nakazuje okenske odprtine na severni steni severne cerkve (zid 1), po eno na prezbiterialnem delu obeh sten osrednje in južne cerkve, ki gledata na vmesni prostor (zidova 6 in 14), najverjetneje na vzhodni in zahodni steni "memorije" (zidova 9 in 10) in najmanj eno okno (v prezbiterialnem delu) v južni steni južne cerkve (zid 13). Zid 13 stoji nad prepadno steno, tako da je nekaj okenskega stekla zelo verjetno zdrselo po pobočju. presenetljivo stekla ni najdenega v prezbiteriju osrednje cerkve, kjer bi ga najbolj pričakovali. Morda sta imeli severna in osrednja cerkev okna le na vrhu zidov daljših stranic, medtem ko so prezbiterialni del razsvetljevale steklene svetilke. Na južni strani je bil na prezbiterij osrednje cerkve prizidan manjši prostor nepravilne štirikotne oblike ("me-morija", glej tudi pogl. 2.5.3 in 4.2.8). Pred prehodom iz osrednje cerkve v omenjeni prostor je v sekundarni legi ležal velik kamnit blok s profiliranimi robovi (pril. 4). Glede na njegovo lego sklepamo, da je služil kot arhitrav na prehodu iz prezbiterija v domnevno memorijo. 3.3.2 ZIDAN PROSTOR MED PREZBITE-RIJEMA OSREDNJE IN JUŽNE CERKVE ("MEMORIJA") Manjši pravokotni prostor je bil prizidan na prezbiterij osrednje cerkve, vendar ni jasno, ali hkrati z gradnjo prvih dveh cerkva ali nekoliko pozneje. Prostor je bil brez estriha, hodno površino je predstavljala delno umetno zravnana skalna osnova, v katero je bila vklesana jama pravokotne oblike (sl. 2.71). V polnilu jame (SE 21) je bilo najdenih nekaj ožje neopredeljivih odlomkov grobe poznoantične keramike in železa. Jama verjetno predstavlja prazen grob - kenotaf. Možnost, da gre za kasneje izropan grob, je izključena zaradi kompaktnega polnila jame in kamnite plošče ter kamnov na vrhu (sl. 2.71). Funkcija prostora ni povsem jasna. Možno je, da je bil v uporabi kot memorija. Precejšnje število drobnih najdb (predvsem steklenih kozarčkov in balzamarijev) dopušča tudi možnost, da je šlo za zakristijo (Tonovcov grad. Najdbe, sl. 3.2). Nekoliko manj verjetno se zdi, da to when the churches were rebuilt. The small amphora with a silver coin of Justinian I in it was placed into a widened crevice that reached under the altar (Fig. 2.65; Tonovcov grad. Finds, Pl. 101: 7) This find is linked to the last rebuilding of the churches, during which the mortar floors were created and the banks and altars were built. These works can be dated to the mid 6th century according to the minting period of the coin (540-552, see also Tonovcov grad. Finds, chapter 5.1, Cat. No. 157) and the dating of the amphora (Tonovcov grad. Finds, chapter 4.1.7). The natural gap in the rock that ran along the longitudinal axis of the main church and was slightly widened under the presbytery partition wall was filled with Antique material and could have been used as the baptistery basin - piscina (see chapter 4.2.6). The partition wall was not preserved at this spot and it is possible that it did not exist at all (Insert 2). The pit of grave 5 located in the northwest corner of the main church cut through the mortar floor of the nave (Insert 2), thus the burial had to take place sometime after the mortar floor was set, i.e. after the mid 6th century. Taking into account the quantity of the discovered window glass fragments it could be assumed that the church windows were glazed (Tonovcov grad. Finds, Fig. 3.3). The position of the glass fragments indicates that windows were located in the north wall of the north church (wall 1), one in both, main and south church presbytery walls overlooking the area between the churches (walls 6 and 14), probably on the east and west sides of the 'memoria' (walls 9 and 10) and at least one window in the south wall (presbytery part) of the south church (wall 13). Wall 13 is positioned above a straight rock face so some window glass could have slid down the slope and was not found. Surprisingly, no window glass was found in the presbytery of the main church where it would be expected. perhaps the north and main churches only had windows on top of the lateral walls while the presbyteries were lighted by glass lamps. A small square space ('memoria', see chapters 2.5.3 and 4.2.8) was attached to the south side of the main church's presbytery. A large stone block with profiled edges was found in a secondary position, in front of the passage from the main church into the square room (Insert 4). Taking into account its position we can assume that it functioned as an architrave between the presbytery and the assumed memoria. 3.3.2 A LATERAL SQUARE SPACE BETWEEN THE PRESBYTERIES OF THE MAIN AND SOUTH CHURCH ('memoria') A small square room was attached to the presbytery of the main church, however it is unclear whether it was bi bil prostor uporabljan kot baptisterij. Tej interpretaciji nasprotuje predvsem dejstvo, da je bil vhod v prostor mogoč samo skozi prezbiterija osrednje in južne cerkve. Kljub temu je mogoče, da je bil prostor kot baptisterij uporabljan v prvi razvojni fazi, torej v času, ko južna cerkev še ni bila sezidana. V tem primeru bi bil možen vhod skozi vrata v zidu 11, krstilni bazenček pa bi bil lahko iz neobstojnega gradiva (lesa; glej tudi pogl. 4.2.6). 3.3.3 JUŽNA CERKEV Na "memorijo" je bila s prezbiterijem prizidana še južna cerkev (pril. 3, 4). Zid 14, ki predstavlja severno stranico ladje južne cerkve, se ne priključi neposredno na vogal zidov 10 in 11, ampak se izteče pribl. 0,5 m južneje od vogala memorije. Nato se pravokotno zalomi in se priključi na vogala zidov 10 in 11 (sl. 2.47, 2.73; pril. 4). Taka nepravilna povezava kaže, da je bila južna cerkev zgrajena pozneje kot memorija, lahko pa si jo razlagamo tudi kot posledico prilagajanja razgibanemu skalnemu terenu. Orientacija južne cerkve odstopa od orientacije severne in osrednje, kar je prav tako lahko posledica prilagajanja skalnemu terenu, mogoče pa je tudi, da je njena usmeritev namerna (glej pogl. 4.3). Tako prezbiterij kot ladja južne cerkve sta bila prevlečena z estrihom, ki je bil nanesen neposredno čez skalno osnovo. ohranjeni so bili le še skromni ostanki ob zidovih. Kamnita pregrada, ki je ločevala prezbiterij od ladje, je bila nad hodnim nivojem ladje ohranjena do višine 30 cm. Grajena je bila kakovostno, iz dokaj velikih kamnitih lomljencev, široka 50 cm, kar je več kot pregradi v severni in osrednji cerkvi, ki sta bili široki med 30 in 40 cm. Stopnice, ki bi vodile iz ladje v prezbiterij, niso bile najdene. Klop za duhovščino v južni cerkvi je bila izdelana precej slabše kot v severni in osrednji (pril. 3). V veliki meri je bila za klop uporabljena kar skala, ki se na tem mestu močno dvigne nad hodni nivo cerkve. Skala je bila obklesana, dozidani so bili le manjkajoči deli. Pri klopi je opazna razlika v izdelavi severne stranice v primerjavi s preostalo klopjo, ni pa jasno, ali razlika v gradnji pomeni tudi kronološko razliko. potem ko je bila klop že izdelana, je bil v ozek prostor med njeno južno stranico in južnim zidom klopi položen skelet, ki je bil na vrhu prekrit s tegula-mi in ometom (grob 1, sl. 2.74, glej tudi pogl. 2.5.4). Verjetno hkrati z grobom je bil med južno stranico klopi in cerkvijo izdelan še pravokoten prostor, zaprt z 1,1 m dolgim zidom, čez katerega so bile nato položene kamnite plošče (sl. 2.75). Tako izoblikovan prostor je bil zapolnjen z ostrorobim gruščem in brez najdb, ki bi omogočile datacijo. Morda gre tudi pri tej strukturi, podobno kot pri "memoriji", za kenotaf. Glede na to, da je bil prečni zidec, ki je zapiral prostor, postavljen čez na added at the same time as the two churches were built or somewhat later. The room did not have a mortar floor, and the floor level was represented by a partially levelled out bedrock into which a rectangular pit was cut (Fig. 2.71). A few fragments of Late Antique coarse ware and a few iron objects were discovered among the pit fill (SU 21), however it is impossible to provide an exact date for these fragments. It is highly likely that the pit represents an empty grave - a cenotaph. The compact fill of the pit and a stone slab and stones on top of it (Fig. 2.71) led us to dismiss the possibility of a robbed grave. The function of this area is not entirely clear. It might have been used as a memoria. The large number of small finds (especially glass vessels and balsamaria) led to the idea that this could have been a sacristy (Tonovcov grad. Finds, Fig. 3.2). It is somewhat less likely that the space was used as a baptistery. This interpretation is most strongly opposed by the fact that the entrance into the room was only possible through the presbyteries of the main and south church. However, it is possible that the room was used as a baptistery during the first development phase, i.e. at the time when the south church did not yet exist. If this was the case the room could have been entered through a door in wall 11, while the baptistery piscina could have been made from non-lasting material (wood; see also chapter 4.2.6). 3.3.3 SOUTH CHURCH The presbytery of the south church was attached to the 'memoria (Inserts 3, 4). Wall 14, which represents the north side of the south church's nave, is not attached to the corner created by walls 10 and 11, but runs approximately 0.5 metres south of the memoria corner. Then it turns at a right angle and joins the corner of walls 10 and 11 (Figs. 2.47, 2.73; Insert 4). This unusual connection indicates that the south church was built after the 'memoria', even though it could also be explained as merely an adjustment to the diverse rocky terrain. The orientation of the south church differs from that of the north and main church. This could also be a result of adjustments made due to the rocky terrain, however it is also possible that its orientation is intentional (see chapter 4.3). The south church's presbytery and nave were covered with a mortar floor that was placed directly onto the bedrock. Only modest remains of this mortar floor were discovered next to the walls. The stone partition wall that separated the presbytery from the nave was preserved up to 30 cm above the floor level of the nave. Its workmanship was of high quality, it was made from large quarry stones and measured 50 cm in width, which is more than the partition walls in the north and main church, which were between 30 and 40 cm wide. No stairs that would lead from the nave into the presbytery were found. tem mestu še ohranjen cerkveni tlak, lahko tako grob kot pravokotni prostor za njim umestimo prav v zadnjo fazo delovanja cerkva, v čas po sredini 6. st. 3.3.4 NARTEKSI Severna in osrednja cerkev sta imeli enoten narteks, prizidan na zidova 4 in 8, ki predstavljata zahodni stranici cerkva. V notranjosti je narteks delila velika (pribl. 1,2 m) višinska razlika med južnim (višjim) in severnim (nižjim) delom. Na višjem delu je hodno površino predstavljala skalna osnova oziroma tanka zemljena plast (SE 31) nad njo. Ostanki estriha niso bili najdeni. V severnem, nižjem delu se je ohranilo več kulturnih plasti, ki jih lahko postavimo v čas obstoja cerkva (sl. 2.68). V plast grušča nad skalno osnovo sta bila vkopana grobova (grob 4 in grob 13). Prav pod grobom 13 je bila najdena tudi noga steklenega kozarčka (sl. 3.24: 11), datirana v konec 5. in 6. st. (glej Tonovcov grad. Najdbe, pogl. 3.3, 3.4; t. 59: 14). Plasti malte in večjih kosov ometa, ki so bile najdene v osrednjem delu narteksa (SE 50, 52), so verjetno ostanek s sten odpadlega zidu in ne estriha. Vprašanje, ali je bil narteks v zadnjem obdobju prekrit z estrihom, tako ostaja odprto. Zaradi nizke višine ohranjenih zidov vhod v narteks ni bil najden. Domnevamo pa ga lahko na zahodnem vogalu severnega zidu (zid 21), kjer je bila v zidu vidna izravnava, ki bi lahko bila osnova za prag. Problematična ostajata tudi prehoda iz nižjega dela narteksa v višji del ter iz narteksa v severno cerkev. Na obeh mestih je bilo namreč treba premostiti precejšnjo višinsko razliko, vendar ni bilo najdenih nobenih ostankov stopnic. Domnevamo lahko, da so bile lesene. Zidovi narteksa so bili grajeni z malto, čeprav je je bilo ohranjene precej manj kakor pri zidovih cerkva. Kaže tudi, da niso bili ometani, saj ob stranicah v notranjosti ni bilo plasti padlega ometa. Edini ostanek ometa na zidu je bil dokumentiran ob grobu 4, kjer je bil ometan del zidu, ki je služil kot stena grobnice (sl. 2.69). Majhna količina ruševinskih kamnov (predvsem na delu ob osrednji cerkvi) kaže, da ni bil pozidan v celotni višini, ampak je bil odprt. ostankov strešne kritine je bilo v ruševini narteksa zelo malo, zato domnevamo, da ni bil krit z opeko. Časa izgradnje narteksa severne in osrednje cerkve ne moremo natančno določiti. Glede na analogije z drugimi najdišči domnevamo, da sta bila zgrajena sredi 6. st., takrat ko so bili v ladjah položeni estrihi in postavljene klopi (glej pogl. 3.3.2). Tudi narteks južne cerkve je bil prizidan na zahodni cerkveni zid. Je zelo nepravilne oblike in slabo ohranjen, na nekaterih mestih tudi popolnoma uničen. Kljub temu je vidno, da je grajen v enaki tehniki kot narteks severne in osrednje cerkve. Tudi tu ni bilo ohranjenih nobenih The clergy bank in the south church was of much poorer workmanship than the ones in the north or main church (Insert 3). To a large extent the rock - which at this point rises sharply above the floor surface of the church - was used as the bank. The rock was carved, and only the missing parts were added. A difference can be noticed between the craftsmanship on the north side of the bank and the rest of it, however it is unclear whether the difference in the construction also represents a chronological difference. Once the bank was finished, a grave, covered by tegulae and plaster (grave 1, Fig. 2.74, see also chapter 2.5.4) was placed into the narrow space between its south side of it and the south wall. At the same time a rectangular space was created between the south side of the bank and the church wall. This space was closed off with a 1.1 m long wall covered by stone slabs (Fig. 2.75). The space was filled with sharp gravel and did not include any finds that would enable precise dating. It is possible that this structure - similar to the 'memoria' - was a cenotaph. Taking into account the fact that the wall that closed off the area was set on the preserved church mortar floor, we can place the grave as well as the rectangular space behind it into the last period in which the church was still in use, into the time after the mid 6th century. 3.3.4 NARTHICES The north and main churches shared a narthex. It was added to walls 4 and 8 that represent the west walls of the churches. In the interior the narthex was split in two by the large difference in height (approximately 1.2 m) between the south (higher) and north (lower) part. In the higher part the floor level was represented by the bedrock or in some places a thin layer of soil (SU 31) that covered it. No remains of a mortar floor were discovered. A number of cultural layers were preserved in the northern (lower) part and they can be dated into the period during which the churches were in use (Fig. 2.68). Two graves were dug into the gravel layer above the bedrock (graves 4 and 13). The glass goblet foot (Fig. 3.24: 11) dated into the end of the 5th or the 6th century was found under grave 13 (Tonovcov grad. Finds, chapters 3.3, 3.4; Pl. 59: 14). The layers of mortar and larger patches of plaster that were found in the central part of the narthex (SU 50, 52) are most likely remains of the plaster that fell off the wall and not remains of a mortar floor. We are therefore still uncertain whether the narthex was covered by a mortar floor during its final period. Due to the low wall remains the entrance into the narthex was not found. We assume that it was located in the west corner of the north wall (wall 21) where a levelled out area was discovered - possibly the base for the threshold. sledov estriha, hodno površino je predstavljala skalna osnova. V ruševini je bila pred začetkom izkopavanj najdena bronasta posoda (Tonovcov grad. Najdbe, t. 50: 1). Na zunanji strani zahodnega zidu narteksa je ležal grob 3 (sl. 2.76), v katerem je bila pokopana 40-60 let stara ženska (glej tudi Tonovcov grad. Najdbe, pogl. 7). Po pridatkih (srebrn prstan, ogrlica in uhana) lahko grobno celoto datiramo v začetek 7. st. ali kasneje (glej tudi Tonovcov grad. Najdbe, pogl. 2.3, t. 51: 4-7). 3.3.5 PROSTOR MED OSREDNJO IN JUŽNO CERKVIJO vglobljen prostor med osrednjo in južno cerkvijo je v osnovi verjetno naravnega nastanka, saj so podobne kotanje vidne tudi drugje po Tonovcovem gradu. Vendar pa stranice, ki se z vseh štirih strani strmo spuščajo v globino, in skoraj vodoravno dno prostora na več mestih kažejo sledove umetne obdelave (glej pogl. 2.5.5, sl. 2.80, 2.82, 2.83). Dno tako izoblikovanega prostora leži pribl. 2 m pod hodnim nivojem cerkva, njegove dimenzije na dnu pa so pribl. 12 x 3,2 m (sl. 2.81, 2.88). Na nekaterih mestih so bili še ohranjeni skromni ostanki malte, s katero je bilo premazano dno (sl. 2.84). Malta je bila zelo slabo ohranjena in včasih komaj ločljiva od skalne osnove. Časovno opredelitev prostora otežuje predvsem dejstvo, da plasti nimajo neposredne povezave s plastmi cerkvenega sklopa. Dodaten problem so recentni vkopi v osrednjem delu prostora, ki so prekinili tudi odnose med plastmi v njegovem severozahodnem in jugovzhodnem delu (glej pogl. 2.5.5). Na dnu objekta, v glineni plasti, ki je zapolnjevala razpoke v skali, je bil najden močno fragmentiran lonček z rdečim premazom (Tonovcov grad. Najdbe, t. 101: 5), ki bi lahko kazal že na zgodnjeantični čas, vendar je bil poleg najden tudi odlomek strešne opeke, kar kaže, da lahko plast prej povežemo s časom obstoja cerkva in je lonček v njej residualna najdba. Tudi v plasteh, ki so ležale neposredno na skalnem dnu, je bilo najdeno gradivo, ki ga lahko povežemo s cerkvami. Tako je bil na dnu ob skalni steni, na kateri je bila postavljena osrednja cerkev, najden razbit cel strešnik. Iz ruševinskih plasti neposredno nad skalnim dnom izvira tudi nekaj odlomkov steklenih posod in svetilk, datiranih v 6. st. (Tonovcov grad. Najdbe, pogl. 3.3.5) Že malo nad dnom objekta so ležale različne plasti malte (sl. 2.81,2.88,2.90). Nekatere so bile mehke, prhke, vsebovale so tudi prst, kamenje in strešnike, druge pa so s trdo, ravno površino spominjale na estrih. Odnosi med njimi so dokaj nejasni (sl. 2.90). Najdbe v teh plasteh so časovno raznolike. Tako lahko glaziran lonček (Tonovcov grad. Najdbe, t. 101: 4), katerega ostanki so bili najdeni v različnih maltnih plasteh, datiramo v konec 4. st. oz. prvo polovico 5. st. (glej tudi Tonovcov grad. Najdbe, The way the lower and the higher part of the narthex were connected and how the access into the north church was managed remains unknown. Here a rather large difference in height had to be overcome, however no remains of stairs have been discovered. It can be assumed that they were made of wood. Mortar was used in the construction of the narthex walls, however much less of it has been preserved than in the church walls. It seems that the walls were not covered in plaster, for not a single layer of fallen off plaster was discovered alongside the walls in the interior. The only remains of plaster on the wall were documented alongside grave 4, where the part of the wall that served as the tomb wall was covered in plaster (Fig. 2.69). The low number of stones belonging to the destruction phase (especially in the part alongside the main church) show that the narthex was not built up to its entire height, but that it remained open on top. Within the narthex ruins tegulae were rare, thus we assumed that they were not used for roofing. It is impossible to say precisely when the narthex of the north and main church was built. Taking into account the analogies with other sites we can assume that this took place in the mid 6th century, at the same time as the mortar floors in the naves were laid and the banks were made (see chapter 3.3.2). The narthex of the south church was attached to the west church wall. It is of an extremely irregular shape and poorly preserved, in some places completely destroyed. However, it is still clear that it was built using the same technique as was used for the narthex of the north and main church. No remains of a mortar floor were preserved here either, and the walking surface was represented by the bedrock. A bronze vessel was found in the ruins prior to the excavations (Tonovcov grad. Finds, Pl. 50: 1). Grave 3 (Fig. 2.76) in which a 40-60 year old woman was buried, lay on the outer side of the western narthex wall (see also Tonovcov grad. Finds, chapter 7). The grave goods (silver finger ring, necklace and two earrings) date the grave into the beginning of the 7th century or even later (see Tonovcov grad. Finds, chapter 2.3, Pl. 51: 4-7). 3.3.5 THE AREA BETWEEN THE MAIN AND SOUTH CHURCHES The deepened area between the main and south churches is most likely of natural origin, for similar basins can be found elsewhere on Tonovcov grad. However the sides that descend steeply on all four sides and the almost horizontal bottom of the area show traces of carving (see chapter 2.5.5, Figs.2.80, 2.82, 2.83). The bottom of this space lies approximately 2 m under the walking surface of the churches and measures approximately 12 x 3.2 m pogl. 4.1.5). Po drugi strani noge steklenih kozarčkov in dele svetilk, najdene v teh plasteh (Tonovcov grad. Najdbe, t. 63: 2-5,20,22-23,25), lahko bolj povežemo s časom delovanja cerkve v 6. st. Tudi datacija prstana z rumenim kamnom je na podlagi značilnosti izdelave predlagana v 6. st. (glej tudi Tonovcov grad. Najdbe, pogl. 2.3, t. 52: 4). Velika količina strešne opeke, najdena v maltnih plasteh in med njimi, prav tako kaže na čas obstoja ali uničenja cerkva. Nad zgornjimi maltnimi plastmi v jugovzhodnem in severozahodnem delu je ležala debela ruševinska plast (sl. 2.81, 2.88). Tudi tu so bile najdbe kronološko zelo premešane. Tako je bila najdena npr. noga močno pro-filirane antične bronaste fibule (Tonovcov grad. Najdbe, t. 52: 7), pa tudi nekaj odlomkov ustij steklenih kozarcev iz 6. st. (Tonovcov grad. Najdbe, t. 62: 14). V plast ruševine je bila vkopana grobna jama groba 21 (sl. 2.95), ki ga glede na prstan z zakovico lahko umestimo od konca 7. stoletja dalje, najverjetneje pa v 8. stoletje. Posebno zanimiv je ta grob zaradi pridane fibule s čebu-ličastimi glavicami, ki je vsaj 400 let starejša od pokopa (Tonovcov grad. Najdbe, pogl. 2.3, t. 53: 10-12). Skupaj z lego ob cerkvah in lego uhanov, ti so bili položeni v usta pokojne, lahko nakazuje obujanje ali ohranjanje antične krščanske tradicije v zgodnjem srednjem veku. Grob je prekrivala plast velikih kamnov in humusa. Postavlja se vprašanje, kdaj in zakaj so obdelali stene in dno skalne kotanje med osrednjo in južno cerkvijo. Glede na najdbe strešne opeke neposredno na dnu prostora lahko sklepamo, da je čas njegovega nastanka povezan s časom nastanka cerkva konec 5. st. Enega od odgovorov na vprašanje o funkciji prostora morda lahko dajo skromni ostanki malte, s katero je bilo premazano dno. Na njihovi podlagi bi lahko sklepali, da je bil prostor na začetku uporabljan kot vodni zbiralnik. vendar tej tezi nasprotuje dejstvo, da v prostoru ni bilo najdenih ostankov rdečkastega vodoodpornega ometa, značilnega za cisterne, s kakršnim je bila premazana tudi notranjost raziskane cisterne na Tonovcovem gradu (glej pogl. 2.6). vse plasti malte, najdene v notranjosti obravnavanega prostora, so namreč rumenkaste barve, podobno kot plasti s sten odpadlega ometa v notranjosti cerkva. Če je bil prostor namenjen za cisterno ali če je imel kakšno drugo namembnost, je moral biti prekrit. Malo verjetno je namreč, da bi lahko v času obstoja cerkva na platoju, ki je bil že sam po sebi prostorsko zelo omejen, prav med cerkvenimi stavbami pričakovali odprto jamo. Različne maltne plasti, ki so bile najdene v prostoru, so bile v preliminarnem poročilu interpretirane kot tlaki slabe kakovosti (Modrijan 2007, 175-177), vendar natančnejša analiza kaže, da ni tako. Proti hipotezi o tlakih govori predvsem slaba kakovost nekaterih maltnih plasti, pa tudi dejstvo, da so se pojavljale na različnih delih prostora na različnih globinah. Danes se kaže kot bolj verjetna razlaga, da gre v bistvu za ruševinske plasti, ki so nastale ob propadu cerkvenih stavb. Dokaj (Figs. 2.81,2.88). In some places modest remains of mortar that covered the floor were discovered (Fig. 2.84). The mortar was extremely poorly preserved and sometimes almost impossible to distinguish from the bedrock. This area is difficult to date as the layers do not have a direct link to the layers in the ecclesiastical complex. An additional problem is represented by recent trenches dug in the central part of the area, which have disturbed the relations between the layers in its northwest and southeast part (see chapter 2.5.5). A severely fragmented small red coated pot (Tonovcov grad. Finds, Pl. 101: 5) was found at the bottom of the area in the clay layer that filled the gaps in the rock. This pot could be dated in the Early Roman period, however a tegula fragment was found alongside it and this indicates that the layer is more likely linked to the time of the churches and that the small pot is a residual find. Material that can be linked to the three churches was also found in the layers that lay directly on the rock bottom. A tegula was found in fragments at the bottom of the rock face upon which the main church was erected. Few fragments of glass lamps and vessels dated to the 6th century were also discovered in the destruction layers directly on the rock bottom (Tonovcov grad. Finds, chapter 3.3.5). Mortar layers were discovered above the floor of the area (Figs. 2.81,2.88,2.90). Some of them were loose and contained soil, stones and roof tiles, while others were reminiscent of a mortar floor as they were hard and levelled. The relations between them are relatively unclear (Fig. 2.90). The finds in these layers come from different periods. A glazed pot (Tonovcov grad. Finds, Pl. 101: 4), the remains of which were found in different mortar layers, can be dated to the end of the 4th or first half of the 5th century (see Tonovcov grad. Finds, chapter 4.1.5). On the other hand, the stemmed goblet and lamp fragments that were found in these layers (Tonovcov grad. Finds, Pl. 63: 2-5,20,22-23,25) can be linked to the period in which the churches were in use, i.e. to the 6th century. On the basis of the characteristic workmanship the ring with the yellow stone can be dated into the 6th century (see Tonovcov grad. Finds, chapter 2.3, Pl. 52: 4). The high numbers of roof tiles discovered in the mortar layers and between them also indicates that these should be dated into the period when the churches were in use or when they were destroyed. A thick destruction layer was discovered above the upper mortar layers in the southeast and northwest part (Figs. 2.81, 2.88). In this area the finds were similarly chronologically mixed. For example, a fragment of an early Roman fibula (Fig. 2.95) was found in the same layer as fragments of glass rims (Tonovcov grad. Finds, t. 62: 14) from the 6th century. The grave pit of grave 21 (Tonovcov grad. Finds, Pl. 53: 10-12) was dug into the destruction layer. Taking into account the ring with a rivet found in it, it can be obsežne trde plasti malte lahko interpretiramo kot dele s sten cerkva odpadlega ometa, ki se je ohranil dokaj nepoškodovan. Tezo o ruševinskem nastanku teh plasti potrjuje tudi velika količina strešne opeke, najdena v vseh plasteh malte in med njimi. Podobno velja omeniti, da so bili npr. deli enega glaziranega lončka (Tonovcov grad. Najdbe, t. 101: 4) najdeni v več različnih maltnih plasteh, kar govori za njihov hkratni nastanek. Tudi vse steklo kaže oblikovno in časovno sorodnost z najdbami stekla v samih cerkvah. Kot kaže zgornji pregled, so najdbe iz vseh plasti prostora med cerkvami dokaj skromne, podobno kot to velja tudi za cerkve same. Tudi struktura najdb (stekleno posodje, malo grobe keramike) kaže podobnost z najdbami iz notranjosti cerkva. O keramiki lahko rečemo, da tudi po zvrsti izstopa iz siceršnjih razmerij na Tonovcovem gradu. Groba keramika je redka; prevladujejo kakovostno izdelane sklede. Več je lončkov in vrčkov, ki sodijo v skupino namizne keramike, zastopana je tudi glazirana keramika. Morda lahko tudi to keramiko, podobno kot velja to za steklo, povežemo z uporabo v liturgične namene. 3.3.6 GROBOVI V cerkvah in njihovi neposredni okolici je bilo odkritih 13 grobov, en kenotaf in en domnevni kenotaf (glej pogl. 2.5). Najštevilčnejša je bila skupina grobov ob vzhodni stranici cerkvenega sklopa. Tam je bilo najdenih 7 v skalo vklesanih grobnih jam z otroškimi grobovi (sl. 2.77-2.79). Vsi so bili brez pridatkov. En grob (grob 5) je bil najden v ladji osrednje cerkve (sl. 2.67). Njegova grobna jama je prebila estrih v ladji, zato ga lahko datiramo v čas po sredini 6. st. Tudi ta grob je bil brez pridatkov. Grob 4 je ležal v narteksu. Imel je zidano grobnico. Za njeno zahodno stranico so uporabili zahodni zid narteksa, vzhodna in južna stranica sta bili zgrajeni iz velikih kamnov in ohranjeni eno vrsto v višino, ostanki severne stranice pa niso bili najdeni (sl. 2.69, 2.70). Moški skelet v njem je bil brez pridatkov. Grob 13 je ležal severno od groba 4 (sl. 2.69). Bil je vkopan v plast grušča, otroški skelet v njem je bil slabo ohranjen in brez pridatkov. Grob 1 je ležal v prezbiteriju južne cerkve, v prostoru, ki je nastal med južno stranico klopi in južno steno cerkve. Na vzhodni in zahodni strani je bil prostor zaprt še z zidovoma (sl. 2.74). V tako dobljen prostor je bil položen skelet, ki je bil pozneje prekrit s strešniki in ometom. Bil je brez pridatkov. Grob 3 je ležal ob zunanji strani zahodnega zidu narteksa južne cerkve. Na zahodni strani je bila grobna jama obdana z nekaj kamni (sl. 2.76). Ženski skelet v grobu je imel pridane srebrn prstan, ogrlico in uhana (Tonovcov grad. Najdbe, t. 51: 4-7). dated sometime from the end of the 7th century onwards, most likely into the 8th century. This grave is especially interesting because of the crossbow fibula that predates the burial by at least 400 years (Tonovcov grad. Finds, Ch. 2.3, Pl. 53: 10-12). In combination with its position in the vicinity of the churches and the fact that the earrings were placed into the mouth of the deceased, this could indicate a revival or preservation of an ancient Christian tradition in the Early Medieval period. The grave was covered by a layer of large stones and humus. We have to ask ourselves when and why were the walls and the bottom of the rock basin between the main and south church carved. Taking into account the finds of the roof tiles found at the bottom of the area we can assume that this area was created at the same time as the churches were built (end of 5th century). One of the possible answers can be provided by the modest mortar remains that covered the bottom of the area. on the basis of these remains we could conclude that the area was at first used as a water cistern. However, this thesis is opposed by the fact that no traces of the red water-resistant plaster, typical for water cisterns, was found in the space - such plaster was, for example, used to cover the interior of the researched water cistern on Tonovcov grad (see chapter 2.6). All mortar layers discovered in the interior of the treated area are yellowish in colour, similar to the layers of the plaster that fell off the walls in the church interiors. If the area functioned as a water cistern or if it had some other purpose, it had to be covered. It is highly unlikely that an open pit would be created on the already small and limited plateau during the period when the churches were in use. In the preliminary report the various layers of mortar were interpreted as poor quality mortar floors (Modrijan 2007, 175-177), however a closer analysis showed that this was not the case. The poor quality of some of the mortar layers as well as the fact that they appeared in various parts and at various depths speak against the hypothesis of mortar floors. Today we believe it is more likely that they are destruction layers that were formed as the church buildings started deteriorating. The rather large layers of compact mortar can be interpreted as parts of plaster that fell of the church walls, and that remained relatively undamaged. This hypothesis was also confirmed by large quantities of roof tiles that were found in all mortar layers as well as in between them. As an example we should mention that parts of one glazed pot (Tonovcov grad. Finds, Pl. 101: 4) were found in a different mortar layers, which lead us to believe that these layers emerged at the same time. Glass finds also belong to the same forms and period as the ones from the churches. As shown by this overview the finds from all of the layers in the area between the churches are rather modest, and the same could be said for the churches Grob 21 je bil vkopan v ruševino, ki je zapolnjevala vglobljen prostor med cerkvami. Ženski skelet je imel ob levi nadlahtnici majhen nožek, na prstu desne roke prstan z zakovico, ob desni podlahtnici fibulo s čebu-lastimi gumbi, v ustih pa preproste uhane iz bronaste žičke (Tonovcov grad. Najdbe, pogl. 2.3, t. 53: 10-12). themselves. The structure of finds from the area between the churches (glass vessels, few coarse ware fragments) is similar to the structure of finds from the church interiors. As far as pottery is concerned we could say that it is statistically differently represented than in other buildings at Tonovcov grad. Coarse ware is rare. Table ware, represented by small pots and jugs, is more numerous, and certain pieces of glazed pottery were also discovered. It is possible that the pottery, similarly to the glass vessels, could be linked to liturgy use. 3.3.6 GRAVES 13 graves, one cenotaph and one assumed cenotaph were found in the churches and their direct vicinity (see chapter 2.5). The group of graves along the east side of the church complex was the largest. In this group 7 grave pits with the children skeletons, were carved into the rock (Figs. 2.77-2.79). No grave goods were found in any of them. One grave (grave 5) was found in the nave of the main church (Fig. 2.67). This grave cut through the mortar floor in the nave, thus it can be dated after the mid 6th century. No grave goods were found in this grave either. Grave 4 was located in the narthex. It had a masonry tomb. For the west side of the tomb they used the west narthex wall. The east and south sides were built of large stones that are preserved one row in height, while no remains were discovered to the north (Figs. 2.69, 2.70). No grave goods were placed alongside the male skeleton. Grave 13 was dug into a layer of gravel north of grave 4 (Fig. 2.69). No grave goods were discovered alongside the poorly preserved child skeleton. Grave 1 was discovered in the presbytery of the south church, in the space between the south side of the bank and the south church wall. To the east and west it was closed off by two walls (Fig. 2.74). The skeleton was placed into the space and covered with roof tiles and plaster. No grave goods were discovered. Grave 3 was found by the outer western wall of the south church's narthex. A few stones closed off the grave pit on its western side (Fig. 2.76). A silver ring, a necklace and two earrings were placed alongside the female skeleton (Tonovcov grad. Finds, Pl. 51: 4-7). Grave 21, containing a female skeleton, was dug into the ruins that filled the area between the main and the south churches. A small knife was placed alongside her left humerus, a crossbow fibula was placed alongside her right ulna, a ring with a rivet was found on one of her fingers on the right hand, and simple earrings and bronze wires were found in her mouth (Tonovcov grad. Finds, chapter 2.3, Pl. 53: 10-12). 3.4 cisterna (vodni zbiralnik) 3.4 water cistern (reservoir) Za gradnjo vodne cisterne so očitno izkoristili prvotno skalno kotanjo na hribu, ki so jo nato na nekaterih mestih po potrebi razširili oziroma poglobili. Zidovi tako delno temeljijo na skali, delno pa na ilovnati osnovi (glej pogl. 2.6). Cisterna je pravokotnega tlorisa, z dvojnimi zidovi, obdaja pa jo še en zid, ki je predstavljal ograjo okrog cisterne (sl. 2.97). Domnevamo lahko, da so najprej zgradili zunanji obodni zid (SE 209, 210, 211, 212). Zid je imel lepo izdelani obe strani, zunanja je bila tudi ometana do nivoja hodne površine. Notranji obodni zid (zidovi SE 213, 214, 215 in 216) je bil vkopan vsaj 1 m niže v ilovnato podlago ter z veliko količino malte povezan z zunanjim delom. Gradnja notranjih zidov je najbrž potekala skoraj hkrati z izdelavo zunanjega obodnega zidu. Sam zunanji obodni zid namreč ni mogel opravljati funkcije cisterne - bil je preslabo temeljen, preplitev in preozek. Z izdelavo notranjega obodnega zidu so pridobili primerno globino in debelino cisterne. Njeno dno je bilo na gl. 409,35 m n.m., širina obeh obodnih zidov skupaj pa je bila med 1 m in 1,20 m. ohranjena višina notranjega obodnega zidu je bila povsod nižja kot ohranjena višina zunanjega. Zato domnevamo, da je bil notranji del cisterne že v osnovi zgrajen nižje. Stopnička, ki je nastala na ta način, je lahko nosila kritino (deske). Stene in dno zbiralnika so bili premazani z vo-doodpornim ometom, kar so dosegli z veliko količino zdrobljene opeke, primešane v malto. omet je zato oranžno-rdeče barve. omet je bil na nekaterih mestih ohranjen do 0,5 m v višino. Dno je delno predstavljala skala, delno pa glina. Na mestih, kjer je bila osnova glinena, so iz 5-10 cm velikih kamnov naredili podlago, ki so jo nato prevlekli z okrog 5 cm debelim grobim ometom, tega pa še z tanko plastjo bolj finega, gladkega ometa (sl. 2.97,2.103,2.104). Na mestih, kjer je bila osnova skalna, je bil omet premazan neposredno čez skalo. Ograja (SE 205, 206, 207, 208) je bila zgrajena nazadnje in z zidovi samega vodnega zbiralnika nima neposredne fizične povezave. Časovnega razpona med The inhabitants used the natural rocky hollow upon the hill for the construction of the water cistern; wherever necessary they just widened or deepened it. Consequently, the foundations of the walls are bedrock and partly also clay (see chapter 2.6). The cistern of a rectangular ground plan was composed of two adjoined walls (inner and outer section) and encircled by another wall, which represented a fence (Fig. 2.97). It is likely that the outer circle wall was built first (SU 209, 210, 211, 212). Both sides had fine facing and the outer wall was also covered in plaster all to the level of the walking surface. The inner circle wall (walls SU 213, 214, 215 and 216) was entrenched at least 1 m lower into the clay foundation and was adjoined with a large quantity of mortar to the outer section. We assume that the construction of the inner circle wall is probably synchronous with the construction of the outer one. The outer circle wall could not have functioned as a water cistern as its foundations were too low, too shallow and too narrow. Construction of the inner circle wall allowed for the achievement of the appropriate depth and thickness of the cistern. The bottom of the cistern was at 409.35 m a.s.l. and the width of both circle walls together was 1-1.2 m. As the inner circle wall was altogether preserved to a lower level than the outer one, it is surmisable that its height was lower right from the start. The step thus formed could be used for roofing (wooden planks). The walls and the bottom of the cistern were covered by a reddish, water resistant mortar, mixed with a large quantity of crumbled bricks. In certain spots the mortar that covered the inner walls of the cistern was preserved 0.5 m high. The bottom of the cistern was represented partially by rock and partially by sterile clay. Wherever the base was represented by clay, a foundation was made from pebble-stones of 5-10 cm in diameter. This foundation was first covered with approximately 5 cm thick coarse mortar pavement, and over this, a covering of a finer grained mortar pavement was made (Figs. 2.97, 2.103, 2.104). In certain spots the mortar pavement covered the bedrock. gradnjo cisterne in zunanjega obodnega zidu ne moremo ugotoviti, verjetno pa je, da je vsaj nekaj časa vodni zbiralnik deloval brez ograje. Na zahodni strani, obrnjeni proti cerkvenemu kompleksu, je bil v ograji 1,20 m širok vhod, med ograjo in zunanjim obodnim zidom pa je bilo približno 0,6 do 0,8 m prostora, kar zadostuje za nemoteno gibanje. Ograja je predstavljala fizično zaščito zbiralnika. obstaja možnost, da je nosila tudi streho, vendar je manj verjetna. ograja je namreč slabo grajena, hkrati pa bi bila streha na ta način zelo velika. Plast malte SE 222, ki se nadaljuje neposredno iz ometa zidu 212, predstavlja hodno površino med tem zidom in ograjo. Če predpostavimo, da je imel zunanji obodni zid nad danes ohranjeno višino še eno ali največ dve vrsti, bi bila potem njegova višina nad domnevno hodno površino približno med 0,8 m in 0,9 m. Ostankov kritine nismo našli. Nekaj majhnih kosov opeke, najdene v ruševini, je tja prišlo iz sosednjih objektov (objekta št. 30 in 31; sl. 1.7). Možnost opečne strehe je tako izključena, glede na analogije z drugimi poznanimi cisternami pa lahko sklepamo na leseno kritino (Knific 1986, 105-107). Najdbe iz cisterne so izredno skromne in večinoma izvirajo iz premešanih plasti. Iz intaktne plasti na dnu strelskega jarka (SE 230) izvirajo koščki ožje neopredelji-ve prazgodovinske keramike in nekaj atipičnih odbitkov. v plasti tik nad prazgodovinsko (SE 229) je bilo najdeno ustje manjše steklene posodice (Tonovcov grad. Najdbe, t. 63: 30). Noga kozarca je bila najdena v SE 218 pod ograjo cisterne (Tonovcov grad. Najdbe, t. 63: 28) v notranjosti cisterne, v žganinski plasti SE 240, ki je ležala nad plastjo odpadlega ometa ob zidu SE 216, je bil najden keramični lonec (Tonovcov grad. Najdbe, t. 106: 12) in nekaj odlomkov drugih posod. Gre za plast, ki je nastala že po opustitvi cisterne v njeni prvotni funkciji. v ruševinski plasti (SE 201) so bili najdeni puščič-na ost s trikotno ploščato konico in tulom, srp in brus (Tonovcov grad. Najdbe, t. 53: 13-15). Zaradi skromnih najdb je začetek izgradnje cisterne težko natančno določiti. Glede na njeno velikost in način gradnje je možno, da je bila zgrajena hkrati s cerkvenim sklopom, torej konec 5. st. Obstaja pa še druga možnost, to je izgradnja ob zadnjem večjem gradbenem posegu na Tonovcovem gradu, ob prezidavi cerkva sredi 6. st. Cisterna je bila ob propadu naselbine v začetku 7. st. opuščena. Ostanki kurjenja nad plastjo ometa na dnu cisterne pa kažejo, da je bila kasneje uporabljena kot začasno zavetišče. Glede na posode, najdene v tej plasti, lahko domnevamo to uporabo v zgodnjem srednjem veku (Tonovcov grad. Najdbe, pogl. 4.2.5, 4.2.7). Vodni zbiralniki so bili v poznoantičnih višinskih naselbinah neobhoden člen v preskrbi z vodo predvsem v primeru obleganja in so večinoma ležali v bližini cerkva, pa tudi drugih večjih poslopij. Glede na to je bila ob odkritju lega velike cisterne presenečenje, saj na The fence (SU 205, 206, 207, 208) was probably the last to be constructed. It has no direct physical connection with the walls of the water cistern itself. The current stage of investigations does not allow for any determination of the time span between the construction of the water cistern and the fence. Nonetheless, the water cistern presumably functioned for some time prior to the construction of the fence. On the western side and facing towards the ecclesiastical complex, a 1.20 m wide entryway was situated within this fence. There is a space of about 0.6 to 0.8 m between the fence and the outer circle wall, which affords enough room to move about freely. The fence could function as a physical protective barrier for the cistern; it might also have carried a roof, however this possibility is less likely due to the poor construction of the fence. The layer of mortar (SU 222) represented the walking surface around cistern. If we reckon that the wall had one or at the most two more levels above the actual hight, then the height of the wall would exceed the surmised level of walking surface by approx. between 0.8 and 0.9m. We did not find remains of roofing. The modest remains of tiles in the destruction layer of the cistern most likely originate from the neighbouring buildings (buildings No. 30 and 31; Fig. 1.7). The possibility that the roofing was made of brick is eliminated. Analogies with other known cisterns lead us to belive that the roofing was wooden (Knific 1986, 105-107). The material finds from the cistern are modest and mainly come from mixed-up layers. The finds from the intact layers are fragments of prehistoric pottery - which are not more closely chronologically classifiable - and some atypical stone flakes from SU 230. In a layer directly above the prehistoric one (SU 229) a rim of a small glass vessel was found (tonovcov grad. Finds, Pl. 63: 30). A glass goblet foot was found also in SU 218 under the fence of the cistern (Tonovcov grad. Finds, Pl. 63: 28). The remains of a ceramic pot (Tonovcov grad. Finds, Pl. 106: 12) and some others vessels were discovered in the layer of charcoal (SU 240) over the fallen-off plaster near the southern wall SU 216. This layer originates from the time when the cistern was no longer in use. In the demolition layer (SU 201) an arrow-head with a triangular point, a sickle and a whetstone were found (Tonovcov grad. Finds, Pl. 53: 13-15). Because of the modest finds in the cistern, it is impossible to precisely define when the cistern was built. Taking into account the size and masonry technique of the cistern, we assume that it was built at the same time as the ecclesiastical complex, at the end of the 5th century. The other possibility is construction in the middle of the 6th century, during the last major construction works at Tonovcov grad. The cistern was abandoned at the time of the abandonment of the settlement, at the beginning of the platoju s cisterno ni bilo poznanih večjih objektov. Po neurju pozimi 2008, ki je izruvalo in polomilo večino dreves na platoju, pa so se zahodno od cisterne razkrili ostanki še ene večprostorne stavbe (objekt št. 30, sl. 1.7; glej tudi pogl. 1.4) Velika cisterna na najvišjem platoju je zadoščala za hranjenje pribl. 50 m3 vode, ni pa bila edina na Tonovcovem gradu. Poleg nje se v reliefu nakazujeta še dva objekta, pri katerih je mogoče - glede na analogije z drugimi že raziskanimi naselbinami - sklepati na isto namembnost. Na zahodni strani sklopa cerkva je v bližini nar-teksov severne in osrednje cerkve manjša enoprostorna stavba velikosti pribl. 5 x 5 m (objekt št. 29), pri kateri je mogoče glede na lego, kvadraten tloris in vglobljeno notranjost sklepati, da gre za vodni zbiralnik (za drugo mogočo interpretacijo objekta glej pogl. 4.2.6). Tik pod vzhodnim zaključkom domnevne četrte cerkve (objekt št. 8) je v manjšem sedlu viden obris enoprostorne stavbe, ki je mnogo nižja in nekoliko poglobljena (objekt št. 7). V njej domnevamo omenjeni cerkvi pripadajoč vodni zbiralnik. 7th century. The rests of a modest fireplace in the layer above the fallen-off plaster show that, after the abandonment, the cistern was used as a temporary shelter. On the basis of the coarse ware from this layer we can assume this use in the Early Middle Ages (tonovcov grad. Finds, chaps. 4.2.5, 4.2.7). Water cisterns in Late Antique hilltop settlements were an indispensable component for the supply of water, especially in the instance of a siege. They were mainly situated near the churches, as well as other larger structures. Consequently, the positioning of the large cistern at tonovcov grad was surprising, because at the beginning of research no others structures were known in the vicinity. After the terrain was cleared from the storm that toppled and snapped a number of trees in the winter of 2008, another building with several rooms was revealed, situated on the very edge of the plateau with the reservoir (building No. 30, Fig. 1.7; see also chapter 1.4). Based on the dimensions of the cistern, a volume of about 50 m' is anticipated. Surely, the researched cistern was not the only one within the settlement. two more structures are distinguishable alongside it in the surface relief; and, based on analogies with other already researched settlements, they presumably served the same purpose. A small, single-room structure, approximately 5 x 5 m, stood on the western side of the ecclesiastical complex, near the narthices of the north and main church (building No. 29). Its positioning, square ground plan and deep interior suggest it was also a water cistern (for other possible interpretations see chapter 4.2.6). Just below the eastern end of the surmised fourth church (building No. 8), the contours of another single-room structure can be traced in the relief (building No. 7). This structure is much lower and somewhat deeper and is presumably the corresponding water cistern to the above-mentioned church. 4. primerjave 4. comparisons 4.1 stavba 1 4.1 Building 1 4.2 Interpretacija cerkvenega sklopa 4.2 Interpretation of the ecclesiastical complex 4.3 Astronomska orientacija cerkva na Tonovcovem 4.3 Astronomical orientation of churches at Tonovcov gradu (saša ČAVAL, Ivan ŠPRAJc) grad (saša ČAVAL, Ivan ŠPRAJc) 4.1 stavba 1 4.1 building 1 skromni ostanki zidov in kulturnih plasti v izkopnem polju stavbe 1 kažejo na obstoj zidane stavbe že v času prve intenzivnejše poselitve hriba, torej v poznorimskem obdobju (faza PA 1). Njena velikost ni znana, ležala pa je približno na istem območju kot poznejša stavba. Analiza najdb v plasteh znotraj in zunaj poznejše hiše ter skromni ostanki zidov so dokazali, da jo je glede na bolje opredeljive najdbe mogoče datirati v drugo polovico 4. in začetek 5. st. (glej pogl. 3.1.3). Njenega propada ne moremo natančneje opredeliti, morda se je to zgodilo še precej pozneje, kot nakazujejo zadnje dobro opredeljive najdbe iz prvih treh desetletij 5. st. Tako se dvoplastnost stavbnih ostalin na poznoantičnih postojankah vzhodnoalpskega območja potrjuje tudi tu: poznamo že nekaj najdišč s poznorimskimi zgradbami, vendar so ta v glavnem slabo raziskana in še slabše objavljena. Časovno sorodne so le preliminarno objavljene stavbe z Brinjeve gore (Pahič 1960; Pahič, Koprivnik 2002, 14-15) kot tudi večidel le v površini razvidna poznorimska arhitektura z Ajdovščine nad Rodikom, kjer pa se nakazujejo veliko kompleksnejše tlorisne zasnove (Slapšak 1997, 50-51). Najbolje so znane zidane in lesene zgradbe z Ančnikovega gradišča pri Jurišni vasi, ki jih je na podlagi drobnih najdb mogoče umestiti v poznorimsko obdobje in so bile obljudene vse do sredine 5. st. (Strmčnik - Gulič, Ciglenečki 2003, 11-21). Tudi pod cerkvami na Ajdovskem gradcu nad Vranjem omenjajo stavbne ostaline starejšega obdobja, ki bi jih glede na odkrite glazirane posode smeli datirati v isto obdobje (Ulbert 1975, 56). Na Invillinu so ugotovili za čas do sredine 5. st. povsem drugačne zidane strukture, ki so jih pozneje nadomestile lesene zgradbe (Bierbrauer 1987, 90-122). obraten primer je bil odkrit na Korinjskem hribu, kjer je ob koncu 5. st. zgrajen vodni zbiralnik presekal ostaline lesene stavbe z gradivom iz 4. in delno 5. st. (Ciglenečki 1985, 256-257). veliko jasnejša je slika, ki jo dajejo v celoti ohranjeni temelji stavbe 1 iz konca 5. in 6. st. (sl. 2.2; pril. 1). Za njeno postavitev so izkoristili najugodnejše mesto na izravnanem in zatišnem delu naselbine (sl. 4.1, 4.2). The modest wall remains and cultural layers within the excavation area of building 1 indicate the existence of a masonry building already during the first more intensive settlement of the hill, i.e. in the Late Roman period (phase LA 1). Its size is unknown, but it stood roughly in the same location as the later building 1. The analysis of the reliable finds in the layers under the later building and the modest wall remains indicate that the building can be dated to the second half of the 4th and the beginning of the 5th century (see chapter 3.1.3). It cannot be precisely ascertained when it was abandoned. This could have happened much later as indicated by the last clearly definable finds from the first three decades of the 5th century. Thus, this is another case to confirm that building remains in Late Antique posts found within the East Alpine area are two-phased. There are quite a few known sites that include Late Roman buildings, however most of them are poorly researched and published. From this period we only have the preliminary publication on the buildings from Brinjeva gora (Pahič 1960; Pahič, Koprivnik 2002, 14-15), and the Late Roman architecture from Ajdovščina above Rodik, which is mainly only visible in the surface, however very complex ground plans have been noticed (Slapšak 1997, 50-51). The best known are the stone and wooden buildings at Ančnikovo gradišče near Jurišna vas, which were dated (according to small finds) into the Late Roman period and that were in use until the mid 5th century (Strmčnik - Gulič, Ciglenečki 2003, 1-21). Older building remains are also mentioned under the churches on Ajdovski gradec above Vranje.Taking into account the discovered glazed ware these remains could be dated into the same period (Ulbert 1975, 56). The settlement at Invillino revealed completely different masonry structures from the period reaching up to the mid 5th century that were later replaced by wooden buildings (Bierbrauer 1987, 90-122). The opposite situation is represented at Korinjski hrib, where the water cistern built at the end of the 5th century cut through the remains of wooden building with finds dating to the 4th and partially 5th century (Ciglenečki 1985, 256-257). The fully preserved foundations of building 1 from the end of the 5th and 6th century at Tonovcov grad offer Sl. 4.1: Tonovcov grad, stavba 1 pred izkopavanjem. pogled z zahoda. Fig. 4.1: Tonovcov grad, building 1 prior to the excavations. A view from the west. Sl. 4.2: Tonovcov grad, stavba 1 po konservaciji. v ozadju zaščitni stavbi nad severno in osrednjo cerkvijo. Fig. 4.2: Tonovcov grad, building 1 after the preservation. protection buildings over the north and main church in the background. Glede na lego stavbe blizu glavnega vhoda (sl. 1.7) bi lahko sklepali na njeno vlogo v zvezi z nadzorom vhodnega območja. Zasnovana je bila kot enoprostorna velika zidana zgradba z vetrolovom in nadstreškom, pozneje so ji dodali manjši, slabše izdelan stranski prostor, ki je bil stisnjen v skromen razpoložljiv prostor na jugozahodni strani. Njena površina (brez nadstreška) znaša 84,7m2. Nastala je v natančno ne povsem opredeljenem času ob koncu 5. st. ali celo na začetku 6. st. (glej pogl. 3.1) in njen tloris je mogoče dobro vzporejati s podobnimi stavbnimi tlorisi na vzhodnoalpskem območju. Začetna enoprostorna zasnova je osnovna oblika stavb v poznoantičnih višinskih naselbinah, saj je bila zaradi omejenega primernega bivalnega prostora in slabših pogojev za gradnjo zidava omejena na najosnovnejše in funkcionalno najprimernejše oblike (pregledno pri Ciglenečki 1987a, 133). Tako ji je glede na dimenzije, izpostavljeno lego kot tudi na pomembne drobne najdbe zelo podobna stavba 2 z Rifnika, ki je merila v začetku 8 x 8,5 m, pozneje pa so jo skrajšali za 1 m (Bolta 1981, 10). Ležala je najvišje izmed bivalnih zgradb, v bližini cerkve in cisterne, v njej in ob njej pa so našli večino langobardske keramike, ki dokazuje, da se je tu nekaj časa zadrževala tudi langobardska družina (Ciglenečki 2001, 193; 2005, 269). Na Ajdni sta bili na spodnji terasi izkopani večji stanovanjski zgradbi (Sagadin 2002-2004, 155-156). V a much clearer image (Fig. 2.2; Insert 1). The building was built on the best possible location, in an area of the settlement that was levelled out and sheltered from the winds (Figs. 4.1, 4.2). Its position next to the main entrance into the settlement (Fig. 1.7) indicates that its function was to monitor the entrance area. In the beginning the building was set up as a large single room stone building with a wind shield and a porch, and later a smaller, poorer quality outhouse was squeezed into the small available space on the southwest side of the house. Its area (without the porch) measured 84.7 m2. The building was erected some time at the end of the 5th or even the beginning of the 6th century (see chapter 3.1) and can be compared to similar ground plans noticed in buildings in the East Alpine area. The single room ground plan represents one of the basic building forms in Late Antique hilltop settlements where the limited space and the poor conditions limited the construction to the most basic and functional forms (see Ciglenečki 1987a, 133). As regards its dimensions, its exposed position and the important small finds this building is extremely similar to building 2 at Rifnik, which at the beginning measured 8 x 8.5 m, but was later shortened by 1 m (Bolta 1981, 10). It was the highest situated house in the settlement, in the vicinity of the church and the cistern. Most of the Lombard pottery found at Sl. 4.3: Ajdna nad Potoki. Konservirani ostanki zidane zgradbe z dobro vidnim vetrolovom na južni strani naselbine. Fig. 4.3: Ajdna above Potoki. Preserved remains of the building with a wind shield on the south side of the settlement. bližini najlažjega dostopa v naselbino je ležala najbolje raziskana stavba, ki je imela ob strani manjši prostor, vhod pa je varoval močan vetrolov (sl. 4.3). stavbi 1 s Tonovcevega gradu je prav tako podobna t. i. stavba B na Ajdovskem gradcu nad Vranjem (Ulbert 1979a, 703-705). Ležala je tik pod cisterno, merila je 9 x 6,5 m in je imela prav tako vetrolov, ki pa je bil usmerjen v njeno notranjost (sl. 4.4). Pri poznejših raziskavah so odkrili tudi večji stranski prostor, ki pa podrobneje ni opisan (Knific 1994, 213, na načrtu stavba 11). veliko bivalno stavbo s prizidkom s Tonovcovega gradu bi smeli primerjati tudi z vsaj dvema pomembnejšima stavbama na Gradcu pri Prapretnem (sl. 5.9), v preliminarni objavi najdišča označenima s številkama 1 in 3 (Ciglenečki 1981, 429, sl. 4). Prva ima povsem primerljivo lego in v vsej dolžini nadzoruje vhod v naselbino. Pri obeh sta opazni v neposredni bližini tudi vodni cisterni in pri stavbi št. 3 še neki manjši dodaten prostor. Rifnik was found within its ruins and around it which proves that a Lombard family spent some time here (Ciglenečki 2001, 193; 2005, 269). Two large houses were discovered on the lower two terraces on Ajdna (Sagadin 2002-2004, 155-156). The best researched one, with a small room added to its side and an entrance protected with a wind shield, stood just next to the easiest access into the settlement (Fig. 4.3). The so-called building B on Ajdovski gradec nad Vranjem is rather similar to the building 1 at Tonovcov grad (Ulbert 1979a, 703-705). It stood just below the water cistern, measured 9 x 6.5 m, and also had a wind shield, only that this one was turned towards the building interior (Fig. 4.4). During later research a large side room was discovered, however it was never described in detail (Knific 1994, 213, building 11 on the plan). The large building with the outhouse at Tonovcov grad could be compared to at least two important build- Sl. 4.4: Ajdovski gradec nad Vranjem. Stavba B z vetrolovom na notranji strani (po Ulbert 1979a, sl. 6). Fig. 4.4: Ajdovski gradec above Vranje. Building B with the wind shield on the inner side (after Ulbert 1979a, Fig. 6). Vendar sta obe zgradbi neraziskani in zato vprašanje vetrolova in nadstreška ostaja odprto. Med že raziskanimi stavbami na slovenskih poznoantičnih najdiščih pa se močno razlikujeta kompleksna zgradba na Kučarju (sl. 4.11) in skupek treh stavb na vrhu Ajdovskega gradca (sl. 4.13). Glede na zasnovo in lego v neposredni bližini cerkva sta bili označeni kot bivališči klera (Petru 1979, 731; Ciglenečki 1995, 185-186). stavbo s podobno namembnostjo domnevamo na Tonovcovem gradu bliže cerkva, kjer se že v površini zarisujeta dve zgradbi z več prostori (objekta št. 28 in 30, sl. 1.7). Na širšem vzhodnoalpskem območju se nakazujejo manjše podobnosti na Sv. Hemi na Koroškem (Glaser 1991, 69-74; Ladstatter 2000, 22-26), kjer so bili pri nekaterih manjših zgradbah (L in M) ugotovljeni sorodni vetrolovi, medtem ko so večje stavbe (H, P, Q) zasnovane večprostorno (sl. 4.5). ings (numbers 1 and 3 as seen in the preliminary publication) at Gradec near Prapretno (Ciglenečki 1981, 429, Fig. 4). The first has a similar position and controlls the entire length of the entrance into the settlement (Fig. 5.9). Both buildings have water cisterns in their close proximity and building No. 3 has an additional small space. However, the two buildings have not yet been researched and thus the issues of the wind shield and porch remain open. Very different from the above described are the complex building on Kučar (Fig. 4.11) and the three buildings at the top of Ajdovski gradec above Vranje (Fig. 4.13). Due to the ground plan and the location (immediately next to the churches) they are believed to have been the living quarters ofthe clergy (Petru 1979, 731; Ciglenečki 1995, 185-186). A building with a similar purpose can be expected to be found at Tonovcov grad, again in the vicinity of the churches, where two buildings with multiple rooms can be seen in the surface (buildings Nos. 28 and 30, Fig. 1.7). Sl. 4.5: Hemmaberg (Sv. Hema). Načrt vzhodnega dela najdišča s cerkvami in stanovanjskimi zgradbami, od katerih imajo nekatere vetrolove (po Glaser 2003, sl. 9). Fig. 4.5: Hemmaberg. Ground plan of the eastern part of the settlement with churches and dwelling houses - some of them with wind shields (afret Glaser 2003, Fig. 9). Dalje proti zahodu so pri izkopavanjih na Monte Barru našli zelo velike bivalne zgradbe, ki kažejo že veliko kompleksnejše tlorise. V prvi fazi so ti sicer imeli preprosto pravokotno zasnovo, vendar so bili v notranjosti večkrat predeljeni. Pozneje so dobili različne predprostore in stranske prostore. posebej podobna stavbi 1 s Tonovcovega gradu je tako stavba 2 (Brogiolo 2001, 24-74). Na veliko manjšem najdišču v bližini, na S. Martino di Lecco, pa so raziskali vsaj dve zelo podobni zidani zgradbi, ki sta bili zaradi velike strmine hriba delno vklesani v skalo. Že od samega začetka sta bili zasnovani kot večji bivalni in manjši stranski prostor z gospodarsko namembnostjo (Brogiolo, Uboldi, Bernardi 2001, 287-298). Če bi v pretres vzeli tudi raziskane lesene stavbe, bi lahko precejšnjo podobnost našli na Tinju nad Loko pri Žusmu (sl. 5.18), kjer med raziskanimi stavbami izstopa predvsem stavba št. 4 z dvema prostoroma in večjim kuriščem zunaj nje. Proporcionalno je podobna stavbi 1 s Tonovcovega gradu, imela pa je prav tako pomembno vlogo v naselbini, na kar kažeta njena velikost in lega na zelo izravnanem, bivalno ugodnem delu pobočja (Ciglenečki 2000, 27-33). Stavbe na Colle Santino v Invillinu, ki sodijo v III. fazo in se časovno ujemajo z obravnavano zgradbo na Tonovcovem gradu, se od nje zelo razlikujejo. Ne samo tehnika gradnje (temelj iz suhega zidu in lesena nadgradnja), ampak tudi zelo podolžna oblika z razmerjem stranic 2 : 1 kažeta njihovo povsem drugačno gradnjo in oblikovanost (Bierbrauer 1987, 294-296). Med njimi po površini (103 m2) izstopa le stavba B, ki bi utegnila imeti v naselbini pomembnejšo vlogo. Skromni zgodnjesrednjeveški ostanki, ugotovljeni v zgornji ruševinski plasti stavbe 1 na Tonovcovem gradu, ne dovoljujejo daljnosežnejših sklepov. Potrjujejo opažanje, da so bile ruševine poznoantičnega objekta preurejene v manjši bivalni (?) prostor, kar nakazuje krajše zadrževanje ljudi na hribu v tem obdobju. Podobne skromne zgodnjesrednjeveške ostaline so bile odkrite tudi v številnih drugih poznoantičnih naselbinah (Ciglenečki 1992a, 56-58; Modrijan 2011). Intenzivno poseljenost nakazuje veliko število kovinskih zgodnjesrednjeveških predmetov na Gradišču nad Bašljem, kjer pa jih doslej ni bilo mogoče povezati s posameznimi bivalnimi objekti (Knific 1999). Številne najdbe potrjujejo umeščenost zgodnjesrednjeveških ostalin v ruševine poznorimskih zgradb na Ančni-kovem gradišču (Strmčnik - Gulič, Ciglenečki 2003, 23-25). Prav tako so bili fragmenti poznoantične keramike pogosto pomešani s posameznimi kosi zgodnjesrednjeveške keramike v nekaterih stavbah na Tinju nad Loko pri Žusmu (Ciglenečki 2000, 153-156). Poseben primer je Sv. Hema, kjer so v vzhodnem kompleksu dvojnih cerkva odkrili zgodnjesrednjeveško keramiko, v kateri vidijo romanske izdelke 7. st., ki In the broader east Alpine area we can notice certain similarities at Hemmaberg in Carinthia (Glaser 1991, 69-74; Ladstatter 2001, 22-26), where similar wind shields have been found in some smaller buildings (L and M), while the other, larger buildings (H, P, Q) had a plan with multiple rooms (Fig. 4.5). Further towards the west the excavations at Monte Barro revealed large buildings with complex ground plans. In the first phase they had a simple rectangular plan, however their interiors were divided by multiple partition walls. Later on various side rooms and anterooms were added. Building 2 is exceptionally similar to the building 1 at Tonovcov grad (Brogiolo 2001, 24-74). At S. Martino di Lecco, a much smaller site nearby, two similar masonry buildings have been researched, both of which were partially carved into the bedrock (due to the steep slope). They were planned as a larger living room and a smaller side room for economic use already from the very beginning (Brogiolo, Uboldi, Bernardi 2001, 287-298). If we also took into account the researched wooden buildings, we could see great similarity also on Tinje above Loka pri Žusmu (Fig. 5.18), where building No. 4 stands out with its two rooms and a large fireplace outside the house. The building has similar proportions as the building 1 from Tonovcov grad, and its size and the extremely favourable position on a levelled out area of the hill indicates that it had an important role in the settlement (Ciglenečki 2000, 27-33). The buildings on colle Santino in Invillino, which belong into phase III and thus to the same period as the discussed building 1 on Tonovcov grad, differ greatly. Not only the building technique (dry wall foundations and wooden superstructure), but also the emphasised prolonged shape with a 2 : 1 ratio of the sides, indicate that a different way of building was implemented (Bierbrauer 1987, 294-296). Only building B stands out with its size (103 m2), thus it is assumed that it could have had an important role within the settlement. The modest Early Medieval remains, discovered in the destruction layers of the building 1 at Tonovcov grad, do not allow for final conclusions. They confirm the observation that a smaller living (?) area was created within the ruins of the Late Antique building, and this indicates that there were shorter periods within which people stayed on this hill. Similar modest Early Medieval remains were discovered in a number of Late Antique settlements (Ciglenečki 1992a, 56-58; Modrijan 2011). Intensive settlement is indicated by the high numbers of Early Medieval metal objects on Gradišče above Bašelj, which could so far not be linked to any individual living quarters (Knific 1999). Numerous Early Medieval finds were also discovered within the ruins of Late Roman buildings on Ančnikovo gradišče (Strmčnik - Gulič, Ciglenečki 2003, 23-25). In some buildings on Tinje above Loka pri Žusmu Late Antique pottery fragments so prišli na to območje po propadu cerkva (Ladstatter 2000, 159-164). Navedene raziskave dokazujejo, da so bile poznoantične naselbine pogosto uporabljene tudi v zgodnjem srednjem veku, vendar zgodnjesrednjeveških bivališč v poznoantičnih naselbinah doslej ni mogoče podrobneje predstaviti: vse kaže, da je šlo za kratkotrajna in zasilna bivališča v največkrat skromno preurejenih ruševinah starejših poznoantičnih objektov. were often mixed with individual fragments of Early Medieval pottery (Ciglenečki 2000, 153-156). A special example is Hemmaberg. There Early Medieval pottery, discovered in the ruins of the eastern double church complex, is believed to be product of the autochthonous people in the 7th century when the churches were no longer in use (Ladstätter 2000, 159-164). Research proves that Late Antique settlements were commonly used during the Early Medieval period, however such Early Medieval settlement remains cannot be presented in greater detail. It seems that they were intended for short stays and most often placed in modestly reworked ruins of older Late Antique buildings. 4.2 interpretacija cerkvenega sklopa 4.2 interpretation of the ecclesiastical complex 4.2.1 OBLIKOVANOST CERKVA Dvoranske cerkve s prosto stoječo klopjo za duhovnike so značilne za vzhodnoalpski prostor in so bile že večkrat obravnavane. O njihovi razprostranjenosti je pred več kot dvajsetimi leti pisal Volker Bierbrauer (Bierbrauer 1988, 45-75) in kljub nekaterim dopolnitvam se slika, ki jo je podal, ni bistveno spremenila. Cerkve s Tonovcovega gradu tako zgolj dopolnjujejo, kar je že dolgo znano. 4.2.1 CHURCH DESIGN Hall churches with a free standing bank for the clergy are characteristic for the east Alpine area and have been discussed on numerous occasions. Over twenty years ago Volker Bierbrauer (Bierbrauer 1988, 45-75) wrote about their distribution and even though certain additions were made, this image of his remains more or less the same. The churches on Tonovcov grad merely add to what was already known. Sl. 4.6: Tonovcov grad. Pogled s severozahodne strani na raziskani sklop cerkva (brez južne cerkve). Fig. 4.6: Tonovcov grad. A view of the ecclesiastical complex (without the south church) from the northwest. Vse tri cerkve na Tonovcovem gradu so po tlorisni zasnovi tipične predstavnice akvilejskega kroga. Imajo preprosto pravokotno zasnovo in v zadnji fazi prislo-njene klopi za duhovnike (sl. 4.6). Značilnosti tovrstnih cerkva so bile podrobneje predstavljene tudi pri obravnavah posameznih najdišč, kot so Ajdovski gradec nad Vranjem (Ulbert 1975, 56-62), Invillino (Bierbrauer 1988, 45-69), Sv. Hema (Glaser 1991, 49-69) in Kučar (Ciglenečki 1995, 134-142). Looking at their ground plans all three churches at Tonovcov grad are typical representatives of the Aquileia circle. They have a simple rectangular plan and the clergy banks were added during the last phase (Fig. 4.6). The characteristics of similar churches were also presented in greater detail in publications of individual sites, such as Ajdovski gradec above Vranje (Ulbert 1975, 56-62), Invillino (Bierbrauer 1988, 45-69), Hemmaberg (Glaser 1991, 49-69) and Kučar (Ciglenečki 1995, 134-142). 4.2.2 PREZBITERIJI IN KLOPI ZA DUHOVNIKE 4.2.2 PRESBYTERIES AND CLERGY BANKS Prezbiteriji vseh treh cerkva so zelo dobro ohranjeni in vsebujejo elemente, ki pomagajo rekonstruirati podobo notranjosti podeželskih cerkva (prim. pri Glaser 2003, 413-420). Dobro je bilo videti vse najpomembnejše sestavne dele kot klopi, katedro, ograje, stopnice, ambon, loculus za relikvije. Prezbiterija v severni in južni cerkvi sta bila prekrita s kakovostnim estrihom, zato se zdi potrebno poudariti, da v osrednji cerkvi ta ni bil najden. Ali gre za njegovo naključno uničenje ali pa ga namenoma niso položili? Pomemben sestavni del prezbiterija v vseh treh cerkvah so zelo dobro ohranjene klopi za duhovnike. Vgrajene so bile v neki poznejši fazi in se naslanjajo na vzhodno steno. so povsem primerljive s tistimi, ki so bile podrobneje obravnavane že pri interpretaciji cerkvenega kompleksa na Ajdovskem gradcu nad Vranjem (Ulbert 1975, 39, 53) in cerkev na Colle di Zuca v Invillinu (Bierbrauer 1988, 50-57) ter na Ku-čarju, kjer smo tudi že preliminarno podali prvi opis klopi za duhovnike s Tonovcovega gradu (Ciglenečki 1995, 134-135). Ne vemo, kakšne so bile klopi v začetni gradbeni fazi. Pomisliti bi smeli na lesene, kot so to domnevali pri zgornji cerkvi na Ajdovskem gradcu nad Vranjem (ulbert 1975, 64), pri starejših cerkvah v Concordiji (iz konca 4. st.) in cerkvi na trgu Piazza Vittoria v Gradežu iz druge polovice 4. st. (Bierbrauer 1988, 77). lesena je bila domnevno tudi klop v manjši cerkvi na Rifniku (Pirkmajer 1994, 48, sl. 59). Podobno bi smeli domnevati tudi pri cerkvi na Ajdni. Morda gre pri vgraditvi klopi za povečanje števila duhovnikov ali celo priselitev škofa, kot sklepa Glaser (1991, 62-63) v primeru povečanja klopi za duhovnike na lavantu. Dobro datirana pozna vgradnja klopi za duhovnike na Tonovcovem gradu nakazuje možnost, da smemo tudi pri drugih sorodnih najdiščih, kjer so znaki večjih dogradenj v cerkvah, domnevati vgradnjo klopi v sredini 6. st. Tako so na obnovo klopi, ki jo je bilo mogoče na podlagi Justinijanovega novca datirati v čas po letu 541, opozarjali že pri raziskavi spodnje cerkve na Vranju, (Ulbert 1979b, 151). The presbyteries of all three churches are well preserved and include elements that help reconstruct the interior of provincial churches (see Glaser 2003, 413-420). All of the important parts, such as banks, the pulpit, fences, stairs, the ambo and the loculus for relics were very well visible. The presbyteries in the north and south churches were covered with a high quality mortar floor, thus it seems important to emphasise that it was not found in the main church. It is unknown whether it was deliberately destroyed or was it intentionally not laid. An important part of the presbyteries in all three churches were the excellently preserved clergy banks. They were built during a later phase and they leaned against the east wall. They can be compared to those that were discussed in greater detail in the interpretations of the church complex at Ajdovski gradec above Vranje (Ulbert 1975, 39, 53) the churches on Colle di Zuca in Invillino (Bierbrauer 1988, 50-57) and on Kučar, where we had also given the first description of the clergy bank at Tonovcov grad (Ciglenečki 1995, 134-135). We are uncertain as to the banks in the preliminary building phase. It is possible that they were wooden, as was assumed for the upper church on Ajdovski gradec above Vranje (Ulbert 1975, 64), the older churches in Concordia (from the end of the 4th century) and the church on the Piazza Vittoria square in Grado (second half of the 4th century; Bierbrauer 1988, 77). It is assumed that the bank in the smaller church on Rifnik was also wooden (Pirkmajer 1994, 48, Fig. 59), and similar could be assumed also for the church on Ajdna. Maybe the new bank was built because the number of clergymen increased or even because the bishop moved in, which was what Glaser assumed was the case in Lavant when the clergy bank was enlarged (Glaser 1991, 62-63). A clearly dated late inclusion of the clergy bank at Tonovcov grad allows us to assume that a mid 6th century bank was added also at other similar sites that reveal signs of later large additions to the church. This idea drew attention to the renovation of the bank at Vranje already during the research of the lower church. The Tudi pri Rifniku vgradnje slabo ohranjene klopi ni bilo mogoče podrobneje datirati, ker so skromni ostanki ležali na goli skali in niso imeli povezave z drugimi deli prezbiterija. Glede na poznejšo dogradnjo baptisterija in narteksa bi smeli domnevati, da je tudi klop nastala šele takrat. Podobno je na Ajdni, kjer je klop evidentno naslonjena na vzhodno steno. Tudi tu ni jasno, ali je nastala ob gradnji cerkve ali pa šele takrat, ko so opustili prvotni narteks in ga prestavili ob vzdolžno stranico cerkve ter vanj verjetno vgradili baptisterij (Sagadin 1989, 248). Zato je možno pri obeh primerjalnih najdiščih vgradnjo njunih klopi datirati podobno kot v primeru Tonovco-vega gradu v čas kratkotrajne bizantinske uprave na tem območju (Ciglenečki 2006, 298-299). Pri osrednji cerkvi na Tonovcovem gradu je treba izpostaviti v zaklinku klopi hrambo steklenih svetilk na manjši, posebej za to prirejeni polici (Ciglenečki 2003, 18). Podoben, podrobneje neopisan prostor je videti na podobnem mestu tudi na severni strani klopi za duhovnike v cerkvi na Ajdni (Sagadin 1989, 249, sl. 79). Pri severni in osrednji cerkvi so iz ladje v prezbiterij vodile stopnice, ki so bile dozidane prav tako v drugi fazi. Sklepati smemo, da so bile v začetku tako kot klopi lesene. Podobne stopnice so našli na Ajdni, in sicer v sredini zidu (Valič 1981, 267). Nenavadna je lega groba 1 in kenotafa v prezbiteriju južne cerkve: moški grob 1 je ležal med južno stranico klopi in južnim zidom cerkve, domnevni kenotaf pa je bil prislonjen na rob klopi za duhovnike. Duhovnik je po Ambroziju pokopan blizu oltarja, kjer je služil (Glaser 1991, 55). 4.2.3 OGRAJE PREZBITERIJEV Pri podeželskih manjših cerkvah na vzhodnoalp-skem območju se je prezbiterij ločeval od prostora za vernike na več načinov. V redkih primerih so bile ograje narejene iz lepo obdelanih stebričev in marmornih plošč, večinoma pa iz lesenih pregrad, kakršnih značilni odtisi so bili odkriti pri spodnji cerkvi na Kučarju in zgornji na Vranju (Glaser 1997, 29-31; Ciglenečki 1995, 135-137). tNa območju province Dalmacije so udi v manjših podeželskih cerkvah za pregrado uporabljali le kakovostnejše, večinoma okrašene kamnite elemente (Chevalier 1995, 129-134). Ločitev prezbiterija od ladje v cerkvah na Tonovcovem gradu močno odstopa od prej navedenih: prezbiterij je bil pri osrednji in še posebej pri severni cerkvi ločen s poudarjeno višjo lego, ki jo je določala predvsem naravna izoblikovanost na vzhodni strani močno dvignjenega terena. Pri vseh treh cerkvah pa so bili prezbiteriji še dodatno ograjeni z manjšim zidcem, na katerem bi smeli domnevati še nižjo leseno pregrado. Značilna je velika višinska razlika med ladjo in prezbiterijem, ki dosega renovation could be dated with the Justinian coin into sometime after 541 Ad (Ulbert 1979b, 151). At Rifnik the inclusion of the poorly preserved bank could also not be precisely dated as the modest remains lay on the bare rock and were not connected to the other parts of the presbytery. Taking into account the later addition of the baptistery and narthex we could assume that the bank was added at the same time. On Ajdna the bank clearly leaned against the east wall. It is unclear whether the bank was added when the church was constructed, or later on, when the primary narthex was abandoned and placed along the longer side ofthe church and a baptistery was included into it (Sagadin 1989, 248). Thus it is possible that the banks were added at both comparable sites at the same time as it was added at Tonovcov grad, i.e. during the short Byzantine presence in this area (Ciglenečki 2006, 298-299). In the main church on Tonovcov grad we have to point out the glass lamps that were stored on a small shelf behind the bank (Ciglenečki 2003, 18). A similar area that has not been described in greater detail was also discovered in the church on Ajdna, where it was also located in a similar place on the north side of the clergy bank (Sagadin 1989, 249, Fig. 79). In the north and main church stairs led from the nave into the presbytery, and it is clear that they were added during the second phase. We can conclude that they were wooden in the first phase, just like the banks. Similar stairs were found on Ajdna, in the middle of the wall (Valič 1981, 267). The positions of grave 1 and the cenotaph in the presbytery of the south church are unusual: male grave 1 was positioned between the south side of the bank and the south church wall, while the assumed cenotaph leaned against the clergy bank. According to Ambrosius the priest was buried close to the altar at which he served (Glaser 1991, 55). 4.2.3 THE PRESBYTERY PARTITION WALLS There were numerous ways of separating the presbytery from the area dedicated to the worshippers in small provincial churches in the east Alpine area. Only rarely were the fences made from nicely carved pillars and marble plates, for mostly they were merely wooden separations, characteristic imprints of which were also discovered in the lower church on Kučar and the upper church on Vranje (Glaser 1997, 29-31; Ciglenečki 1995, 135-137). In the small provincial churches in the province of Dalmatia they used only high quality stone elements that were usually decorated (Chevalier 1995, 129-134). In the churches on Tonovcov grad the presbyteries were separated from the naves in a very different way: in the main and even more so in the north church the pres- pri severni cerkvi kar 60 cm. Le nekoliko manjša je na Ajdni (Valič 1981, 267). 4.2.4 OLTARJI v vzhodnoalpskem prostoru so bili v uporabi predvsem oltarji v obliki miz z enim stebričkom ali štirimi stebrički, zidani so izjema (Ulbert 1988, 290-291; Glaser 1997, 31). Na Tonovcovem gradu so vsi trije oltarji zidani, pri čemer je bil najbolje ohranjen tisti v severni cerkvi, kjer smo v njegovi ruševini odkrili tudi velik odlomek kamnite žare. Ta je bila sekundarno uporabljena kot posoda za relikviarij in so jo po opustitvi cerkve pustili v ruševini. Podobno zidano gradnjo oltarja poznamo tudi z Ajdne (sl. 4.7), kjer so v njegovi neposredni bližini našli tudi bronaste in koščene dele relikviarija (Leben, Valič 1978, 538, 539). pri tej obliki oltarja gre nedvomno za preproste oltarje, ki so bili doslej na vzhodnoalpskem območju znani le z obeh omenjenih najdišč (Glaser 2003, 424). Sodijo tako med prve v vrsti oltarjev, ki so se uveljavili pozneje v zgodnjem srednjem veku (Sennhauser 2003c, 962). Elementov starejših oltarjev nismo našli, zato domnevamo, da so bili v začetni fazi cerkva leseni. Podobno domnevajo pri mnogih drugih cerkvah, npr. pri bytery was separated with its higher position that was a result of the terrain which rises towards the east. In all three churches the presbyteries were additionally separated by a low wall, upon which we can assume stood a low wooden division. The large difference in height between the nave and the presbytery, which measures as much as 60 cm in the north church, is characteristic for the churches on Tonovcov grad. This difference in height is only slightly smaller in the church on Ajdna (Valič 1981, 267). 4.2.4 ALTARS Altars shaped like tables with one column or four columns were characteristic for the east Alpine area; masonry altars were an exception (Ulbert 1988, 290-291; Glaser 1997, 31). All three altars on Tonovcov grad are built with stone and the best preserved is the one in the north church, where a large fragment of a stone urn was discovered in its ruins. This was secondarily used as a vessel for the reliquary and when the church was abandoned it was left in the ruins. A similar masonry altar was found on Ajdna (Fig. 4.7), where also bronze and bone parts of a reliquary were found in its vicinity (Leben, Valič 1978, 538, 539). Sl. 4.7: Ajdna nad Potoki. Ostanki zidanega oltarja. Fig. 4.7: Ajdna above Potoki. Remains of the mansonry altar. zgornji cerkvi na Vranju, pri Laubendorfu idr. (Ulbert 1975, 64; Glaser 1997, 130; Sennhauser 2003c, 961). Prav zaradi dobro datirane prenove prezbiterija na Tonovcovem gradu z Justinijanovim novcem je mogoče tovrstne oltarje povezati z bizantinsko prenovo v času Justinijanove rekonkviste. podobno velja tudi za oltar cerkve na Ajdni, ki je prav tako lahko nastal v tej obliki šele ob prenovi cerkve in ne že v začetku kot sestavni del prvotne zgradbe. V skrbno položenem spatejonu v vdolbini, ki je vodila pod oltar, bi smeli domnevati olje ali kakšno drugo organsko snov, ki je bila prej v dotiku z relikvijami in je zato veljala za "dotaknjeno relikvijo" (prim. pri Glaser 1997, 53). 4.2.5 AMBON Je v cerkvah akvilejskega kroga redko ohranjen. Dobro je viden na Lavantu, kjer je na pravokotni osnovi okrogel nastavek (Miltner 1953, 62, Glaser 1997, 143). Podoben podaljšek je v Invillinu, a ga Bierbrauer zaradi prisotnosti domnevne soleje ni označil kot ambon, čeprav dopušča to možnost (Bierbrauer 1988, 59). Pred raziskavami na Tonovcovem gradu smo ga lahko na slovenskem območju še najzanesljiveje potrdili na Ko-rinjskem hribu, kjer je kot večji umetno obdelan kamen ležal tik pred prezbiterijem (Ciglenečki 1985, 257). Na sosednjem dalmatinskem območju so amboni datirani v pozno obdobje obstoja cerkva in kažejo na vpliv Ravene ali Konstantinopla (Chevalier 1995, 153). Pozni fazi je mogoče pripisati tudi ambon na Tonovcovem gradu: bil je evidentno prizidan ob pregrado prezbiterija severne cerkve in njegovo vgradnjo bi smeli datirati v čas največje prenove cerkva na tem najdišču. Podobno pozno datacijo bi smeli domnevati tudi na Korinjskem hribu, kjer se zdi verjetno, da so prav ob justinijanski prenovi cerkve dodali tudi ambon. 4.2.6 BAPTISTERIJ Poseben problem je identifikacija baptisterija oziroma njegove piscine v cerkvenem sklopu. S. Ciglenečki je v začetni fazi raziskav na Tonovcovem gradu ugotavljal piscino v umetno razširjeni skalni razpoki pred prezbiterijem osrednje cerkve predvsem zaradi ustrezne velikosti jame, pa tudi domnevno pripadajočega zravnanega prostora v prezbiteriju tik nad jamo (Ciglenečki 1997b, 21). F. Glaser je v osrednji cerkvi (B) na Tonovcovem gradu domneval krstilnico in v severni cerkvi (A) cerkev za maševanje, podobno kot tudi na Sv. Hemi v zahodni dvojni cerkvi (Glaser 1997, 87). Ob enaki velikosti cerkva na alpskem območju je menil, da je bila cerkev s piscino uporabljena tudi za birmanje, tako kot v zahodni dvojni cerkvi na Sv. Hemi (Glaser 1997, 43). Undoubtedly, these are simple altars that have in the east Alpine area only been found in the two aforementioned sites (Glaser 2003, 424). They are therefore one of the first in the line of masonry altars that became popular in the Early Medieval times (Sennhauser 2003c, 962). We had not found any elements of older altars, thus we assumed that they were wooden in the first phase of the church. similar is also assumed for numerous other churches, e.g. the upper church on Vranje, the church at laubendorf, etc. (Ulbert 1975, 64; Glaser 1997, 130; Sennhauser 2003c, 961). Due to the precisely dated renovation of the presbytery at Tonovcov grad (with Justinian's coin) such altars can be linked to the Byzantine renovation during the period of Justinian's Reconquista. Similar holds true for the altar in the church on Ajdna, which was also made in this form only once the church was renovated and not when the original building was erected. In the carefully placed spatheion in the niche under the altar, we could assume oil or some other organic matter that was in contact with relics and was thus believed to be a 'touched relic' (see Glaser 1997, 53). 4.2.5 AMBO Ambones are rarely preserved in the churches belonging to the Aquileian circle. It is clearly visible at Lavant, where a round structure on a rectangular base was found (Miltner 1953, 62, Glaser 1997, 143). A similar extension was discovered in Invillino, but due to the presence of an assumed solea Bierbrauer did not categorise it as an ambo, even though he did not reject this possibility (Bierbrauer 1988, 59). Prior to the research on Tonovcov grad an ambo could be most reliably confirmed on Korinjski hrib (within the territory of present day Slovenia), where it lay in front of the presbytery as a larger man-worked stone (Ciglenečki 1985, 257). In the neighbouring Dalmatia ambones are dated towards the end of the churches' lifespan and they show the influence of Ravenna or Constantinople (Chevalier 1995, 153). The ambo on Tonovcov grad can also be placed into a later phase: it was clearly added to the presbytery partition wall in the north church during the largest church renovation at this site. A similarly late date could also be assumed on Korinjski hrib, where it seems likely that the ambo was added during the Justinian renovation of the church. 4.2.6 BAPTISTERY Identifying the baptistery or its piscina within the ecclesiastical complex has proven to be an especially difficult task. During the first phase of the research on Tonovcov grad S. Ciglenečki ascertained that the piscina was located v osrednji cerkvi ugotovljena domnevna piscina je nenavadna in delno tudi vprašljiva, zato zahteva nekaj pojasnil. Poglejmo argumente za njen obstoj. Zdi se, da je bil nastanek domnevne piscine v celoti pogojen z naravno skalno špranjo, ki se pojavlja na tem mestu. Ustreza glede velikosti in globine. v prid njeni namembnosti govori umetno izravnan manjši prostor velikosti 70 x 30 cm na prezbiterialni strani, tik ob robu prezbiterija, ki nakazuje mesto, ki ga je uporabljal episkop ali prezbiter pri obredu krsta. Izravnan prostor na tem mestu je sicer brez pomena, funkcioniral je lahko samo v povezavi z dogajanjem v cerkveni ladji oziroma bližnjem vglobljenem prostoru. Morda bi tovrstno funkcijo potrjevala tudi koncentracija steklenih predmetov v bližini umetno zravnanega prostora v prezbiteriju (tonovcov grad. Najdbe, sl. 3.2), saj so bile tovrstne posode včasih uporabljane pri krstu. Posredno krstni funkciji objekta pritrjuje ambon v manjši, severni cerkvi, ki kaže, da so jo vsaj v drugi fazi uporabljali kot cerkev za maševanje. Lega domnevne piscine je zelo nenavadna in na vzhodnoalpskem območju nima veliko primerjav. Še najbližja ji je krstilnica v prezbiteriju severne stavbe zahodne dvojne cerkve na Sv. Hemi (sl. 4.8), kjer pa F. Glaser domneva cerkev, namenjeno zgolj birmanju in krščenju (Glaser 1997, 118). Posebej zanimiva je lega dveh piscin v poznoantičnih kastelih v Srbiji, v Boljetinu in Babotincu (Ilič 2006, 226, 229). Ker gre za preproste cerkvene stavbe, je pomembno, da sta piscini ležali v prezbiteriju, in sicer na njegovem robu. Predstavljata primera, ko so bili zaradi pomanjkanja prostora prisiljeni napraviti baptisterij kar v prezbiteriju, kar bi utegnilo biti razlog za nenavadno lego tudi na tonovcovem gradu. Opredelitvi skalne razpoke kot piscine baptisterija pa oporeka nekaj drugih okoliščin. Nenavadna je umestitev krstilnice v osrednji cerkvi, ki je največja in je zato najverjetneje služila maševanju v veliki naselbini. omenili smo že samo lego piscine, ki je kljub nekaterim zgoraj naštetim analogijam v alpskem prostoru nenavadna in izjemno redka. Posebej vprašljiva se zdi prisotnost klopi za duhovnike vsaj v drugi fazi, ki je razumljiva samo, če je cerkev hkrati služila tudi v druge namene. v domnevni piscini stene niso bile ometane, na dnu nad sterilno ilovico ni bilo opaziti estriha. Manjkajo tudi v skalo vsekane stopnice, čeprav bi seveda smeli dopustiti obstoj lesenih. Zasutje jame je vsebovalo časovno podrobneje neopredeljivo poznoantično gradivo, kar kaže, da je bila zasuta v zadnjem obdobju funkcioniranja cerkve. Ob navedenih pomislekih je treba opozoriti tudi na druge možne lokacije baptisterija. tako bi za to funkcijo hipotetično prišla v poštev stavba št. 2. ta bi glede na lego v neposredni bližini cerkva, velikost in orientacijo kot tudi na potek vzhodne stene v podaljšku vzhodnih sten cerkva prav tako lahko predstavljala baptisterij. to within the additionally widened crack in the rock in front of the presbytery of the main church. This conclusion was based on the size of the pit, and the assumed manmade area in the presbytery above the pit (Ciglenečki 1997b, 21). F. Glaser assumed a baptistery in the main church (B) at Tonovcev grad; he also assumed that the north church (A) was intended for mass, similar as was the case with the western double church on Hemmaberg (Glaser 1997, 87). As the churches in the Alpine area were roughly of the same size he assumed that the church with the piscina was also used for confirmation, similar as in the western double church at Hemmaberg (Glaser 1997, 43). The piscina that was ascertained in the main church is unusual and questionable, thus it demands a few explanations. Let's look at the arguments for its existence. It seems that the assumed piscina was conditioned entirely by the natural gap in the rock. The gap is fitting as regards its size and depth. The small manmade levelled out area measuring 70 x 30 cm alongside the presbytery, which indicated the spot that was used by the bishop or presbyter during baptism, also supports this theory. The levelled out area does not have any purpose, and it could function only in relation to the events taking place in the church nave or in the hollowed area nearby. Additional conformation of this function could be found in the concentration of glass objects in the vicinity of the levelled out area in the presbytery (Tonovcov grad. Finds, Fig. 3.2), for such vessels were also used for baptising. The baptism function is indirectly confirmed also by the ambo in the smaller, north church, which indicates that it was used as a church for mass in its second phase (if not also before). The location of the assumed piscina is unusual for the eastern Alpine area and does not have a lot of comparisons. The closest comparison can be found in the baptistery in the presbytery of the north building of the western double church at Hemmaberg (Fig. 4.8), for which F. Glaser assumed that it was intended solely for confirmations and baptisms (Glaser 1997, 118). Especially interesting is the position of two piscinae in Late Antique settlements in Serbia, in Boljetin and Babotinac (Ilič 2006, 226, 229). Because these are simple church buildings it is important that the piscinae were placed on the edge of the presbyteries. These are two examples in which the lack of space forced the builders to build the baptistery within the presbytery, and we think that a similar reason could be behind the unusual position of the baptistery at Tonovcov grad. Some circumstances oppose defining the rock crevice as the baptistery piscina. It is unusual for the baptistery to be placed in the main church, which was the largest and thus most likely served for mass within the large settlement. We have already mentioned the position of the piscina, which is - regardless of the previously mentioned analogies in bi nakazovali posredno morda tudi štirje grobovi ob njeni južni stranici (sl. 2.44). Vendar je bila stavba slabo grajena in ohranjena, odkrili nismo nikakršnih znakov piscine, čeprav tudi znakov, ki bi opozarjali na drugačno, npr. stanovanjsko uporabo, ni bilo. Tudi vseh grobov ni mogoče zanesljivo datirati, enega pa po pridatkih lahko postavimo celo v zgodnji srednji vek (glej Tonovcov grad. Najdbe, pogl. 2.2). Naslednja možna lokacija bi bil prizidek osrednje cerkve, v katerem sicer domnevamo memorijo (glej pogl. 3.3.2). To bi bilo mogoče predvsem, če je bila piscina manjša in prenosljiva. Tudi odkriti stekleni predmeti v severovzhodnem delu temu ne nasprotujejo, kot tudi ne kenotaf ob južni steni. Seveda pa bi bila tovrstna funkcija prostora mogoča le pred izgradnjo južne cerkve, ko je bil vanj mogoč dostop z južne strani. Kot zadnja možnost se ponuja še neraziskana manjša zgradba v osi severne cerkve (objekt št. 29, sl. 1.7), ki pa je od cerkva oddaljena že pribl. 15 m. ob sedanjem stanju raziskav ostaja vprašanje bap-tisterija odprto, še najverjetnejša se zdi njegova lokacija v začetni fazi pred prezbiterijem osrednje cerkve, a tudi te ni mogoče zagovarjati v celoti. 4.2.7 NARTEKSI Narteksi na Tonovcovem gradu so bili prizidani na že obstoječe cerkvene stavbe. To je sicer značilnost velikega števila cerkva na vzhodnoalpskem območju. Med bližnjimi omenimo le Rifnik, Korinjski hrib, Ajdno, Vranje in zgornjo cerkev na Kučarju. Njihove izgradnje ni mogoče povsod zanesljivo datirati: vsekakor so nasta- Sl. 4.8: Hemmaberg (Sv. Hema). Severna cerkev zahodnega cerkvenega sklopa z baptisterijem v prezbiterialnem delu (Glaser 1993, sl. 5). Fig. 4.8: Hemmaberg. North church of the western ecclesiastical complex with baptistery in the presbitery (Glaser 1993, Fig. 5). the Alpine space - unusual and extremely rare. The presence of the clergy bank during the second phase seems especially questionable, for it could only be understood if the church was also used for other purposes. The walls in the assumed piscina were not covered in plaster, and no mortar floor was laid on top of the sterile clay. No stairs seemed to have been cut into the rock, however we should allow for the possibility of wooden ones. The pit fill included Late Antique material that could not be precisely dated, however these finds indicate that it was filled while the church was still in use. Alongside the stated reservations we should also draw attention to the other possible locations for the baptistery. Theoretically, building No. 2 could also be used to perform this function. taking into account its position in the proximity of the churches, its size and orientation, as well as the extension of the churches' east walls, this could be used as a baptistery. In a way, this could be indicated also by the four graves alongside its south side (Fig. 2.44). However, the building was poorly built and thus poorly preserved, and no signs of a piscina have been discovered, even though there are no signs that would indicate any other use (for instance living quarters). Not all of the graves could be precisely dated, and according to the grave goods found in one of them, one could be dated as late as the Early Middle Ages (see tonovcov grad. Finds, chapter 2.2). The next possible location could be the rectangular room between the main and the south church, which we assumed was used as a memoria (see chapter 3.3.2). This could be the case if the piscina was small and mobile. The glass objects discovered in the northeast part do not oppose this idea, neither does the cenotaph found alongside the south wall. Of course, this function would only be possible before the south church was built, when it could still be entered from the south. As the last possibility we have to consider the still un-researched small building in the axis of the north church (building No. 29, Fig. 1.7), however, this building is approximately 15 metres away from the churches. At the moment the question as regards the location of the baptistery remains open, however it seems most likely that it was at first located in front of the presbytery of the main church (which cannot be proven). 4.2.7 NARTHICES On Tonovcov grad the narthices were added to existing church buildings, which was characteristic of li, potem ko so bile cerkve že nekaj časa v uporabi. Večjo sorodnost s Tonovcovim gradom kažeta le dogradnji na Rifniku in Korinjskem hribu, kjer so dodali hkrati baptisterij in narteks in bi ti dogradnji smeli datirati prav tako v sredino 6. st. (Ciglenečki 2006, 298-299). Drugače je na Kučarju, kjer je bil narteks glede na okvirno datacijo celote prizidan že v nekem zgodnejšem obdobju (Ciglenečki 1995, 152). 4.2.8 PROSTOR MED OSREDNJO IN JUŽNO CERKVIJO ("MEMORIJA") Stranskim prostorom cerkva ni mogoče vedno zanesljivo določiti pomena, posebej če manjkajo značilni elementi (Bierbrauer 1988, 58-59; Sennhauser 2003c, 957-961). To na Tonovcovem gradu velja v kar največji meri za pravokoten prostor, ki je bil na južni strani prizidan k osrednji cerkvi. Sprva je bil ta prostor označen kot memorija oziroma kot prostor s kenotafom pomembnega prebivalca naselbine ali duhovnika (Ciglenečki 1997d, 19-20; 2003, 16). Prav tako ga je označil tudi F. Glaser v svojem pregledu zgodnjekrščanskih cerkva v alpskem prostoru ("Gedächtniskapelle"; Glaser 1997, 87). Glede na to, da niso bili najdeni zanesljivi elementi, ki bi v celoti potrjevali takšno namembnost, je treba še enkrat kritično pregledati vse možnosti. Raziskovalci tovrstne prostore največkrat interpretirajo kot zakristijo (Sennhauser 2003c, 957-961). V tem primeru se zdi ta domneva izključena, ker je bil prostor posebej označen s profiliranim arhitravom, kar kaže na njegov večji pomen. Prav tako je treba omeniti, da zakristije niso bile vedno prisotne, saj so liturgične priprave marsikje hranili kar ob klopeh za duhovnike ali kje drugje (Sennhauser 2003b, 24). Prav to smo ugotovili tudi v osrednji cerkvi, kjer je bilo na polici za klopjo za duhovnike shranjenih kar nekaj steklenih kozarcev in svetilk. Druga možnost je grobnica, ki je prav tako zelo pogosta v stranskih prostorih cerkva (Sennhauser 2003b, 26-38). Takšne poznamo med bližnjimi cerkvami v prizidkih na Ajdni kot tudi v zgornji cerkvi na Vranju (Valič 1981, 266; Ulbert 1975, 29-30). Za funkcijo grobnice se zdi nenavadna predvsem velika prazna površina v prostoru, zato ni mogoče povsem izključiti tudi uporabe lesene mize. Tretja možnost je memorialna namembnost prostora. Prisotnost obdelanega nadvratnika, ki je v cerkvenem kompleksu edini okrašeni del stavbne opreme, kaže, da je šlo za neko posebno pomembno mesto v sklopu cerkva, v katerega je vodil glavni vhod prav iz prezbiterija osrednje cerkve. R. Sennhauser poudarja, da je za potrditev funkcije memorije potrebna prisotnost relikvij (Sennhauser 2003b, 34-36). Nanje bi v tem prostoru morda kazali odlomki več steklenih kozarcev z nogo in brez nje ter manjših stekleničk, v katerih so numerous churches in the east Alpine area. Amongst the ones in the vicinity we should mention Rifnik, Korinjski hrib, Ajdna, Vranje and the upper church on Kučar. Their construction cannot be precisely dated: however, we know for sure that they were built once the churches were already in use. Greater similarity with Tonovcov grad is shown merely by the later additions on Rifnik and Korinjski hrib, where the baptisteries and narthices were added at the same time. These additions could be dated into the mid 6th century (Ciglenečki 2006, 298-299). Taking into account the dating of the churches the narthex on Kučar was added somewhat earlier (Ciglenečki 1995, 152). 4.2.8 THE AREA BETWEEN THE MAIN AND SOUTH CHURCH {'MEMORIA') The rooms found at the sides of churches cannot always be ascribed an exact function, especially if the typical elements are missing (Bierbrauer 1988, 58-59; Sennhauser 2003c, 957-961). On Tonovcov grad this holds true for the rectangular area that was added to the south wall of the main church. At first this area was considered to be a memoria or a space with a cenotaph dedicated to an important inhabitant of the settlement or a priest (Ciglenečki 1997d, 19-20; 2003, 16). In his overview of the Early Christian churches in the Alpine area F. Glaser also thought it had this function ('Gedächtniskapelle'; Glaser 1997, 87). Taking into account that no reliable elements were found that would undoubtedly confirm such use, we have to critically review all possibilities. Researchers usually interpret such spaces as the vestry (Sennhauser 2003c, 957-961). In this case this assumption does not seem appropriate, as the area was marked with a profiled architrave, which indicates its importance. We should also mention that the vestry was not always present, as the liturgy objects were often kept alongside the clergy banks or elsewhere (Sennhauser 2003b, 24), which was the case in the main church, where the shelf behind the clergy bank held quite a few glass vessels and glass lamps. The second possibility is a tomb, which was also common in the side rooms in churches (Sennhauser 2003b, 26-38). Such tombs were found in the nearby church on Ajdna as well as in the upper church on Vranje (Valič 1981, 266; Ulbert 1975, 29-30). It seems unusual that merely a large empty area would be used for the tomb, thus the use of a wooden table cannot be excluded. The third possibility is that the area was used as a memoria. The presence of the ornamented architrave, which is the only decorated part of the architectural fittings in the entire ecclesiastical complex, indicates that this was a special place, the main entrance into which lead from the presbytery of the main church. R. Sennhauser emphasised that relics would need to be found if we najverjetneje hranili olje, vino ali kadilo za liturgične namene. Ni mogoče izključiti tudi možnosti, da so v teh stekleničkah hranili olje, ki je bilo v dotiku z relikvijami. Prav tako v prvi fazi, ko še ni bilo južne cerkve, ni mogoče izključiti funkcije baptisterija. Seveda pa v tem primeru preseneča nadvratnik, ki je posebej označeval vhod s strani prezbiterija osrednje cerkve. Tovrstna funkcija je bila torej mogoča le v primeru, da so nad-vratnik tja vstavili pozneje, za kar pa manjkajo dokazi. Nedvomno pa je bila ta hipotetična funkcija prostora spremenjena, ko je bila zgrajena južna cerkev in je tudi drugi vhod v prostor vodil iz prezbiterija južne cerkve, kamor katehumeni niso smeli. Sklenemo lahko z ugotovitvijo, da predvsem posebej označen vhod ob siceršnjem popolnem pomanjkanju okrasnih kamnitih elementov, velik prazen prostor, kjer je morda stal lesen oltar in v bližini morebitne relikvije v steklenih posodah, v kar največji meri kažejo na funkcijo memorije, čeprav tega z vso zanesljivostjo ne moremo potrditi. 4.2.9 PROSTOR MED OSREDNJO IN JUŽNO CERKVIJO POD MEMORIJO wanted to confirm that this area was used as a memoria (Sennhauser 2003b, 34-36). The best indication of this can be found in the fragments of glass beakers or goblets and small bottles that were most likely used to store oil, wine or incense for liturgy purposes. Of course, we should not exclude the possibility that the small bottles were used to store oil that had been in touch with the relics. In the first phase, when the south church was not there, it is also possible that it was used as a baptistery. Of course, if this is the case, the architrave cannot be explained, for it marked the entrance from the presbytery of the main church. In order for us to assume that this used to be the baptistery, we would need to believe that the architrave was a later addition, however there is no evidence to support this. This hypothetical function would have changed when the south church was built and the second entrance was made from the presbytery of the south church, to which the catechumen had no access. We can conclude with the ascertainment that the marked entrance amongst the otherwise undecorated stone elements, a large empty room, where a wooden altar might have stood, and the possible relics in glass vessels, all indicate that this area could have been used as a memoria, even though it is impossible to reliably confirm this. Delno umetno obdelanega prostora prav tako ni mogoče zanesljivo opredeliti zaradi pomanjkanja značilnih elementov in zahtevne terenske konfiguracije. Zelo verjetno smemo v prvi fazi domnevati obstoj cisterne (pogl. 3.3.5). Podobno skromno izdelan estrih poznamo pri cisterni v prvotni cerkvi na Sv. Hemi, kjer malti ni bila primešana opeka (Glaser 1991, 67). Številne so primerjave za bolje zidane cisterne ob cerkvah, ki možnost obstoja preproste cisterne dodatno potrjujejo (prim. pri Chevalier 1995, 95; Gunjača 2005, 39-42). Kako je bil ta prostor oblikovan pozneje, ni bilo mogoče ugotoviti, zdi se, da je bila na tem mestu najverjetneje prekrita vdolbina, v kateri so se pozneje ob propadu cerkva kopičile ruševinske plasti. 4.2.10 GROBOVI Grobovi v cerkvi in ob njej so značilni za zgodnje-krščansko miselnost in izražajo željo biti pokopan ad sanctos (Glaser 1997, 52-53; Sennhauser 2003b, 26-38). Številni grobovi, odkriti v cerkvah in ob njih na Tonov-covem gradu, kažejo zanimivo strukturo pokopanih, saj je mogoče zaslediti kar 8 otroških, 3 moške in 2 ženska skeleta ter en zanesljiv in še en demnevni kenotaf. Osredotočanje grobov ob vzhodni steni osrednje cerkve nakazuje pomembnost tamkajšnjih relikvij. Morda bi smeli ob domnevno "dotaknjenih" relikvijah v amfori domnevati tudi relikviarij, ki je bil lahko - kot kaže del preostale urne v severni cerkvi in primerjava z Ajdno - shranjen v lokulu v odprtini oltarja. Pri večini 4.2.9 THE AREA BETWEEN THE MAIN AND SOUTH CHURCH BELOW THE 'MEMORIA' The partially manmade area could not be reliably defined due to the lack of typical elements and the demanding terrain configuration. We can assume a cistern in the first phase (chapter 3.3.5). A similarly modest mortar floor was found in the cistern in the original church at Hemmaberg, where no bricks were mixed into the mortar (Glaser 1991, 67). There are numerous comparisons for better built cisterns next to churches, which additionally confirm the possibility for the existence of a simple cistern (see Chevalier 1995, 95; Gunjaca 2005, 39-42). It is impossible to ascertain how this area was shaped in the later times, but it seems that there was a covered space in which the destruction layers gathered when the churches started falling apart. 4.2.10 GRAVES The graves in the church and alongside it are typical of the Early Christian mentality and express the desire to be buried adsanctos (Glaser 1997, 52-53; Sennhauser 2003b, 26-38). On Tonovcov grad, the numerous graves discovered in the churches and alongside them show an interesting structure of the buried: so far 8 children, 3 male and 2 female skeletons have been discovered as well as one certain and once assumed cenotaph. bližnjih najdišč so bili grobovi odkriti v narteksih cerkva (Invillino, Korinjski hrib, Rifnik, Vranje). Zanimivo je, da so grobovi najštevilčnejši prav na Ajdni, ki tudi sicer kaže največ podobnosti pri cerkvenih elementih (Sagadin 1989). od navedenih grobov na Tonovcovem gradu izstopata domnevni kenotaf in pokop v bližini klopi južne cerkve (grob 1). 4.2.11 KRONOLOGIJA Kronološki okvir zgodnjekrščanskih cerkva je pri večini težko začrtati zgolj na podlagi značilnih najdb v njih, ker so te redke in največkrat neizpovedne. Raziskave na Tonovcovem gradu pa so vendarle dale kar nekaj bolje datiranih elementov, ki v grobem določajo časovni okvir od konca 5. st. do začetka 7. st. in predvsem pomembno dodelavo cerkva sredi 6. st. (glej pogl. 3.3). Seveda je datacijo cerkva mogoče podati posredno tudi glede na datacijo celotne naselbine, saj so bile cerkve znotraj domišljene naselbinske zasnove njihov integralni del. Bolje opredeljeni predmeti in novci potrjujejo ta kronološki okvir. Tej dataciji pritrjujejo tudi druge bolje raziskane poznoantične utrjene naselbine, pri katerih je bilo mogoče kronološki okvir določiti posredno s pomočjo pripadajočih raziskanih grobišč. Tako je na Rifniku razpon določen vse od konca 5. do konca 6. st. (Bolta 1981, 17), v Kranju datiran v čas med 500 in 600 (Vinski 1980, 18-19), v Teurniji v zadnjih šest desetletij 6. st. (Piccottini 1976, 115) in na Sv. Hemi okvirno v 5. in 6. st. (Glaser 1985, 85; Ladstatter 2000, 24). Omogočajo grobo potrditev začrtanega časovnega orisa in pritrjujejo začetku izgradnje naselbine v konec 5. st., njeno opustitev pa v konec 6. in najverjetneje še v začetek 7. st. posebej pomembna se zdi potrjena datacija prenove in dogradnje cerkva v sredini 6. st., ki je doslej na drugih najdiščih ni bilo mogoče opredeliti z zanesljivo datiranimi konteksti. Podane so bile le približne datacije na podlagi podatkov iz zgodovinskih virov, predvsem z omembo kratkotrajne bizantinske prevlade. Na Tonov-covem gradu pa sta amfora in malo uporabljen novec v njej omogočila povsem zanesljivo datacijo postavitve oltarjev, kakršna sta v severni in južni cerkvi povezana z estrihom, ki prekriva novozgrajene elemente notranje opreme. Tem dogradnjam smemo pripisati zelo verjetno tudi dogradnjo narteksov, ki je sicer časovno podrobneje ni mogoče opredeliti, vendar je zelo verjetno potekala sočasno z drugimi velikimi spremembami v prezbiteri-alnih delih cerkva. Dodatno potrjuje takšno časovno opredelitev tudi z Justinijanovim novcem dobro datirana prenova estriha, ki je prekril tudi klop za duhovščino v spodnji cerkvi na Vranju (Ulbert 1975, 39-41; 1979b, 151; Ciglenečki 2006, 299). The concentration of the graves along the east side of the main church indicates the importance of the relics held here. Apart from the assumed 'touched' relics in the amphora it might be also possible to assume the existence of a reliquary, which could - as shown by the part of the remaining urn in the north church and the comparison with Ajdna - be stored in the loculus in the altar opening. Most of the sites in the vicinity revealed graves within the narthex (Invillino, Korinjski hrib, Rifnik, Vranje). It should be noted that Ajdna has the most graves, and at the same time also shows the most similarities in church elements (Sagadin 1989). From the graves on tonovcov grad the assumed cenotaph and the burial in the vicinity of the bank in the south church stand out (grave 1). 4.2.11 CHRONOLOGY In most cases it is difficult to outline the chronological frame of the Early Christian churches merely with the aid of the typical finds, because these are rare and most often do not reveal much. The research on Tonovcov grad provided a few elements that could be dated and roughly defined the time scale between the end of the 5th to the beginning of the 7th century; it also defined - and this is of utmost importance - the church remodelling into the mid 6th century (see chapter 3.3). Of course, the church can be indirectly dated also in relation to the settlement, for the churches were an integral and well thought out part of the settlement plan. The better defined objects and coins confirm this chronological framework. This date is also confirmed by other better researched Late Antique settlements, for which the chronological framework could be ascertained through the researched graves. for instance, at Rifnik the time span ranges between the end of the 5th to the end of the 6th century (Bolta 1981, 17), in kranj between the years 500 and 600 AD (Vinski 1980, 18-19), in Teurnia in the last six decades of the 6th century (Piccottini 1976, 115) and on Hemmaberg approximately into the 5 th and 6 th century (Glaser 1985, 85; Ladstatter 2000, 24). These examples enable a rough confirmation of the outlined chronology and confirm that the settlement was started at the end of the 5th century, and abandoned at the end of the 6 th or most likely in the beginning of the 7th century. Especially important seems the confirmed dating of the renovation and additions to the churches in the mid 6th century, which has so far not been able to be confirmed with reliably dated contexts at other sites. For other sites merely approximate dates have been given and even these were given on the basis of the data from historic sources, especially the mention of the short Byzantine rule. On Tonovcov grad the amphora and the slightly used coin found within it led to the precise dating odlično analogijo ponuja tudi velika prenova dvojne cerkve v Saloni, kjer monogrami škofa Honorija dopuščajo datacijo nove križne cerkve v drugo četrtino 6. st. (Cambi 2005, 66). opustitev cerkva je mogoče posredno vezati na propad oziroma opustitev celotne naselbine. Zanesljive najdbe segajo do konca 6. st., grob ob stranici narteksa južne cerkve pa s svojimi pridatki dopušča možnost pokopa še v začetku 7. st., kar bi posredno kazalo tudi na obstoj cerkve v tem času. Seveda pa groba, vkopanega v ruševino ob osrednji cerkvi, po pridatkih datiranega v 8. st. (grob 21, glej Tonovcov grad. Najdbe, pogl. 2.3.7), ne moremo vezati na obstoj cerkva v tem času. Tudi improvizirano kurišče v notranjosti osrednje cerkve na tlaku kaže, da so bile takrat že opuščene. 4.2.12 POMEN CERKVENEGA SKLOPA Nenavadno oblikovan skupek treh cerkva na To-novcovem gradu je posebnost med višinskimi naselbinami vzhodnoalpskega območja. Preproste pravokotne zgradbe so povezane v zaključen sklop, ki se je organsko razraščal po razgibanem skalnem površju postojanke in se močno razlikuje od večinoma posamično postavljenih cerkva, delno pa tudi že od dobro znanih dvojnih cerkva. Nadlokalno vlogo naselbine s pomembnim cerkvenim sklopom na Tonovcovem gradu nakazuje dejstvo, da so bila v bližini odkrita kar tri sočasna pribežališča oz. naselbine, ki s svojo majhno površino in skromnim drobnim gradivom kažejo podrejeno vlogo (Gradec pri Logjeh, Gradec pri Drežnici in Sv. Helena v Podbeli, glej pogl. 1.5). Ta najdišča tako potrjujejo poseljenost zaledja Tonovcovega gradu in dovoljujejo domnevo, da je okoliško prebivalstvo v njegovem sakralnem sklopu videlo tudi svoje cerkveno središče. Po pregledu posameznih cerkva in elementov v njih je treba osvetliti funkcijo in pomen celotnega sklopa: cerkve v njem imajo vse značilnosti akvilejskega kroga, v osrednjem delu katerega tudi ležijo, hkrati pa kompleks glede na število zgradb in oblikovanost spominja na nekatere sklope v provinci Dalmaciji. Najprej je treba poudariti, da sta bili severna in osrednja cerkev oblikovani vse od začetka kot dvojna cerkev, ki so jo pozneje le dopolnjevali. Številni grobovi ob in v njej pričajo o pomembnosti relikvij. Ob kakovostni izvedbi posameznih elementov cerkva in opečnatem kritju pa na drugi strani presenečata skromnost notranje opreme in predvsem pomanjkanje njenih okrašenih kamnitih delov, slikanega ometa ter mozaika. Podrobnejšo označbo funkcij posameznih stavb je treba zaradi ne povsem zanesljivo ugotovljenega bapti-sterija pustiti odprto. Pomembno se zdi, da je osrednja cerkev največja in bi zato poleg eventualne krstne funkcije smeli domnevati še kakšno drugo. of the altars, which are in the north and south church linked to the mortar floors that cover the new interior fittings. Even though they cannot be precisely dated the narthices were most likely added at the same time as the other large changes to the presbyteries were made. These dates are additionally confirmed by the Justinian's coin that precisely dated the renovation of the mortar floor which covered the clergy bank in the lower church on Vranje (Ulbert 1975, 39-41; 1979b, 151; Ciglenečki 2006, 299). An excellent analogy is also offered by the great renovation of the double church in Salona, where the monograms of bishop Honorius date the new double church into the second quarter of the 6th century (Cambi 2005, 66). The abandonment of the churches can be linked to the abandonment of the settlements. Reliable finds reach into the end of the 6th century, while the grave goods in the grave along the narthex of the south church allow for the possibility that the burial took place in the beginning of the 7th century, which would indicate that the church was still in use at the time. Of course the grave, dug into the ruins alongside the main church, which was dated with the use of grave goods into the 8th century (grave 21, see Tonovcov grad. Finds, chapter 2.3.7) cannot be linked to the other graves. The improvised fireplace discovered on the paving in the interior of the main church shows that the churches were already abandoned at that time. 4.2.12 THE IMPORTANCE OF THE ECCLESIASTICAL COMPLEX The unusually formed complex of the three churches on Tonovcov grad is a peculiarity amongst the hilltop settlements in the east Alpine area. The simple rectangular buildings are joined into a complex that spread organically along the diverse rocky surface and thus they differ from the individual churches, as well as from the well-known double churches. The more than merely local role of the settlement with the important church complex is indicated by the three contemporary refuges or settlements that were discovered in the vicinity, all of which show a subsidiary role with their small surface area and modest small finds (Gradec near Logje, Gradec near Drežnica and Sv. Helena in Podbela, see chapter 1.5). These sites confirm a settlement in the backdrop of Tonovcov grad and allow for the assumption that the surrounding inhabitants considered this church complex to be their ecclesiastical centre. Following the overview of the individual churches and elements within them we need to shed some light on the function and importance of the entire complex: the churches within it have all of the characteristics of Tri raziskane cerkve predstavljajo zaključen celovit sklop: domnevna četrta cerkev je od ostalih treh nekoliko oddaljena in še v celoti neraziskana, zato jo v nadaljnjo obravnavo ne bomo pritegnili. Za boljše razumevanje sakralnega sklopa pa je treba znova osvetliti fenomen dvojnih cerkva v višinskih naselbinah vzhod-noalpskega prostora in podrobneje raziskati njihovo vlogo v specifični naselbinski podobi tega območja. V doslej najobsežnejši sintezi fenomena dvojnih cerkva je J. P. Sodini posebno pozornost posvetil problemu njihove namembnosti (Sodini 1984, 255-326). Pokazal je na razmerja med različnimi namembnostmi dvojnih cerkva: 41 % jih sodi k škofovskim središčem, 41 % farnim ali naselbinskim cerkvam, 7 % samostanom, 4 % pokopališkim kompleksom in 5,5 % romarskim potem (Sodini 1984, 306). Zato ugotavlja, da dvojne cerkve niso izraz istega programa, ki bi bil veljaven za vse. Posebej poudarja, da vse dvojne cerkve niso škofovske, in obratno, da škofovske cerkve niso vedno dvojne cerkve. Tej problematiki je bil posvečen tudi simpozij v Grenoblu leta 1994, ki je s kar nekaj prispevki osvetlil vzhodnoalpsko območje (predvsem Bratož 1996; Glaser 1996a; Cantino Wataghin 1996). V njih so pregledno predstavljene do takrat znane dvojne cerkve, vendar pa so obravnavane izolirano od poselitvene podobe in strukture spremljajočih zgradb v naselbinah. Glede problematike funkcij in pomena dvojnih cerkva je R. Sennhauser nakazal številne možnosti pri interpretaciji posameznih zgradb, ki jih nikakor ni mogoče enotno tolmačiti (Sennhauser 2003c, 930-933). Na vzhodnoalpskem območju je bilo predvsem v utrjenih višinskih naselbinah odkritih nekaj primerov dvojnih cerkva, ob katerih je tudi posebej zasnovana spremljevalna bivalna in reprezentančna arhitektura, namenjena kleru. Zato se zdi smiselno podrobneje obravnavati prav ta aspekt dvojnih cerkva, ki je največkrat zapostavljen, ter preučiti njihovo vlogo v tedanji poselitveni podobi. Na Tonovcovem gradu pripadajoče zgradbe za kler še niso raziskane. Z veliko verjetnostjo pa bi jih smeli iskati le nekoliko nižje pod cerkvami, kjer se v reliefu kaže izstopajoča večja stavba s petimi prostori (stavba št. 28, sl. 1.7). Prav tako bi glede na bližino prišla v poštev stavba, ki se nakazuje na platoju severovzhodno od cerkvenega sklopa, v neposredni bližini vodnega zbiralnika (stavba št. 30, sl. 1.7). V nadaljevanju bomo pregledali pomembnejše podeželske komplekse dvojnih cerkva v vzhodnoalp-skem prostoru. Pri tem bomo posebno pozornost posvetili tistim dvojnim cerkvam v utrjenih podeželskih naselbinah, kjer so raziskali tudi cerkvenim sklopom pripadajoče zgradbe klera. Na območju zahodno od Tonovcovega gradu se kot analogija za dvojno cerkev nakazujeta predvsem dve najdišči v Karniji. Prvo je že nekaj desetletij raziskan skupek dvoranske cerkve in trichore z baptisterijem na the Aquileia circle, which is appropriate as they are positioned in its central part; at the same time the complex is also similar to certain complexes in the province of Dalmatia - especially as regards the number of buildings and shape. At first we have to emphasise that the north and main church were formed at the very beginning as a double church to which additions were added later on. The importance of the relics is shown by the numerous graves alongside and in the churches. Alongside the high quality of the individual elements and the tegulae, the modesty of the architectural fittings and especially the lack of decorated stone parts, painted plaster and mosaics come as a great surprise. As the baptistery has not been reliably ascertained yet a detailed description of the functions of the individual buildings should be left open. It should be stressed that the main church was the largest and thus it could be assumed that it was not used solely for the eventual baptisms. The three researched churches represent a closed complex: the assumed fourth church is somewhat removed and has not been researched yet, thus it will not be added in the following discussion. For a better understanding of the ecclesiastical complex some light needs to be shed on the phenomenon of double churches and their role in hilltop settlements in the east Alpine area. In the so far most comprehensive analysis of the double church phenomenon, J. P. Sodini paid special attention to their function (Sodini 1984, 255-326). He ascertained the ratios of the various functions of double churches: 41 % of them were a part of episcopal centres, 41 % of them were parish or settlement churches, 7 % belonged to monasteries, 4 % to graveyards and 5.5 % represented a stop on pilgrimage routes (Sodini 1984, 306). Thus he ascertained that not all double churches had the same function and they most certainly did not belong to the same programme. He emphasised that not all double churches were episcopal and vice versa, not all episcopal churches were double churches. The 1994 symposium in Grenoble focused on this issue, and quite a few papers shed some light on the east Alpine area (especially Bratoz 1996; Glaser 1996a; Cantino Wataghin 1996). These contributions included a clear overview of the double churches known at the time; however they were interpreted separately from the settlement image and the structure of the accompanying buildings within the settlements. As regards the function and role of the double churches, R. Sennhauser indicated numerous possibilities when interpreting individual buildings and they can by no means be explained in a unified manner (Senn-hauser 2003c, 930-933). In the east Alpine area there are quite a few double churches, especially in fortified hilltop settlements. These churches were usually accompanied by living quarters and representational architecture, both intended for the Colle di Zuca v Invillinu (Bierbrauer 1988; sl. 4.9). Posebej značilna je lega na manjšem hribcu v bližini naravno veliko bolje zavarovane poznoantične naselbine na Colle Santino (Bierbrauer 1987). Velikost, oprema in lega kažejo na lokalno versko središče, ki je bilo namenjeno ne samo bližnji utrjeni naselbini, ampak verjetno tudi drugim manjšim bližnjim zaselkom. Dvojno cerkev, ki ima značaj cerkvenega središča, kaže drugi cerkveni sklop, ki je bil pred nedavnim od-kopan v bližini Ovara in po kasnejši cerkvi poimenovan San Martino (Cagnana 2008, 448-451). tu so ugotovili dve dvoranski cerkvi in velik oktogonalni baptisterij (sl. 4.10). Cerkvi so postavili na mestu, kjer v bližini ni sočasne naselbine. Prav tako kot na tonovcovem gradu tudi tu v ruševinah niso odkrili veliko okrašenih gradbenih elementov. A. Cagnana domneva podeželsko sakralno središče, ki so ga uporabljali prebivalci iz širše okolice (Cagnana 2008, 448). Najznačilnejši in najbolj celovito raziskan tip sakralnega središča na vzhodnoalpskem območju je bil odkrit na Kučarju pri Podzemlju v Beli krajini (J. Dular, Ciglenečki, A. Dular 1995). Pri sistematičnih raziskavah so bili odkriti dve cerkvi, baptisterij, velika Sl. 4.9: Invillino. Cerkveni sklop na Colle di Zuca (Bierbrauer 1988, sl. 4). Fig. 4.9: Invillino. Ecclesiastical complex at Colle di Zuca (Bierbrauer 1988, Fig. 4). clergy. Thus it seems reasonable to study their role in the contemporary settlement pattern and treat this neglected aspect of double churches in greater detail. On Tonovcov grad the buildings for the clergy have not been researched yet. It is likely that they would be located slightly below the churches, where we can see a large building with five rooms standing out from the relief (building No. 28, Fig. 1.7). Taking into account its proximity, we also have to consider the building shown on the plateau northeast of the church complex, next to the water reservoir (building No. 30, Fig. 1.7). In the continuation we will take a look at the more important provincial double churches in the east Alpine area. At this we will pay special attention to those double churches in fortified provincial settlements where they researched not only the church complexes but also the buildings intended for the clergy that accompanied them. In the area west of Tonovcov grad there is an analogy to the double church in the two sites in Carnia. The first is the complex with a hall church and a trichora with a baptistery at Colle di Zuca in Invillino, which was researched a few decades ago (Bierbrauer 1988; Fig. 4.9). Especially characteristic is the position on a smaller hill in the vicinity of a naturally much better protected Late Antique settlement on Colle Santino (Bierbrauer 1987). Its size, fittings and position indicate a local ecclesiastical centre that was not intended merely for the inhabitants of the nearby fortified settlement, but most likely also for the people living in the small hamlets nearby. A double church (named San Martino after a later church) with the character of an ecclesiastical centre has been indicated by the recently excavated second church complex in the vicinity of Ovaro (Cagnana 2008, 448451). Two hall churches and a large octagonal baptistery were discovered at this site (Fig. 4.10). The churches were erected on a location at which no contemporary settlement was known. Similar to Tonovcov grad the ruins did not reveal a lot of decorated architectural elements. A. Cagnana assumes that this was a provincial ecclesiastical centre, used by the inhabitants of a wider area (Cagnana 2008, 448). The most typical and best researched type of ecclesiastical centre in the east Alpine area was discovered at Kučar near Podzemelj in the Bela Krajina region (J. Dular, Ciglenečki, A. Dular 1995). Systematic research revealed two churches, a baptistery, a large and a small building with living quarters and walls with towers. Taking into account the large empty space in the interior of the fortified settlement we could assume that the inhabitants from the unprotected settlements in Sl. 4.10: San Martino di Ovaro. Sklop cerkva (Cagnana 2008, 449). Fig. 4.10: San Martino di Ovaro. Ecclesiastical complex (Cagnana 2008, 449). in manjša stanovanjska zgradba ter obzidje s stolpi. Glede na velik prazen prostor v notranjosti utrjene naselbine bi smeli sklepati, da so ga prebivalci iz nezavarovanih bližnjih naselbin ob nevarnosti uporabljali tudi kot pribežališče (sl. 4.11). Glede na razmerje med velikostjo cerkva ter pripadajočim bivalnim delom na Kučarju naselja ni mogoče interpretirali kot samostan. Verniki so na Kučar prihajali iz bližnjih zaselkov, do katerih so se lahko vrnili še isti dan. Vlogo Kučarja je tako mogoče razumeti kot sakralno središče manjše ozemeljske enote, ki je ležala proč od pomembnejših cest (Ciglenečki 1995, 189-190). Velika bivalna stavba na Kučarju po velikosti in kompleksnosti tlorisne zasnove presega večino raziskanih bivalnih zgradb v poznoantičnih višinskih naselbinah. Zaradi kratkotrajne uporabe ali pa morda tudi, ker the vicinity used it as a refuge in times of danger (Fig. 4.11). Considering the relation between the size of the churches and the accompanying living quarters the settlement on Kučar cannot be interpreted as a monastery. The churchgoers came to Kučar from the nearby hamlets, to which they could return in the same day. The role of Kučar could thus be understood as an ecclesiastical centre of a smaller territorial unit that was located far from the important roads (Ciglenečki 1995, 189-190). As regards its size and complexity of the ground plan the large building with the living quarters on Kučar is larger than most researched living quarters in Late Antique hilltop settlements. Due to the short period it was in use or maybe also because it was abandoned in times of peace, it did not reveal sufficient small finds that would adequately explain the use of individual rooms. Sl. 4.11: Kučar pri Podzemlju. Sakralno središče iz 5. st. (Ciglenečki 1995, si. 173). Fig. 4.11: Kučar near Podzemelj. Ecclesiastical centre from the 5th century (Ciglenečki 1995, Fig. 173). je bila opuščena v miru, v njej ni dovolj drobnih najdb, ki bi zadovoljivo pojasnile namembnost njenih posameznih prostorov. V kompleksnem tlorisu je izstopala predvsem velika ogrevana soba v močnem nasprotju z ostalimi manjšimi bivalnimi prostori. Veliko bolj izpovedna je bila sicer manjša, a z najdbami bogata cerkvena in bivalna arhitektura, ugotovljena pri raziskavah na Ajdovskem gradcu nad Vranjem (Ulbert 1975; 1979a; Knific 1994). Tu izstopa skupina treh hiš v neposredni bližini dvojne cerkve, ki jih je zaradi enake usmeritve in majhnega razmaka med njimi (0,5 m in 3 m) mogoče presojati kot zaključeno Within the complex ground plan the large heated room stood out when compared to other small living quarters. Much more revealing was the smaller church and living quarters ascertained during the research on Ajdovski gradec above Vranje, both of which were much richer in finds (Ulbert 1975; 1979a; Knific 1994). The group of three houses located next to the double church stands out, and due to their shared orientation and the small distances between them (0.5 m and 3 m) they appear to function as a whole. taking into account the proximity of the churches we can assume that this complex was closely linked to them. These three buildings were the only ones Sl. 4.12: Ajdovski gradec nad Vranjem. Sakralno središče (po Knific 1994, si. 4). Fig. 4.12: Ajdovski gradec above Vranje. Ecclesiastical centre (after Knific 1994, Fig. 4). celoto. Glede na bližino cerkva smemo sklepati na tesno povezavo z njimi (sl. 4.12, 4.13). Te tri zgradbe so edine stale na temenu hriba in imele tako poudarjen pomen Sl. 4.13: Ajdovski gradec nad Vranjem. Prikaz povezav stavbe D z bližnjimi zgradbami (Ciglenečki 1995, sl. 177). Fig. 4.13: Ajdovski gradec above Vranje. Connections between building D and nearby buildings (Ciglenečki 1995, Fig. 177). on top of the hill and had an emphasised role within the settlement (Figs. 4.12, 4.13). The houses have been fully researched and most of the finds have been published (Petru 1979; Knific 1979; 1986; 1994, 212, 214-216; Vičič 1983, 71; Mirnik - Prezelj 1984). The archaeological evaluation shows that the revealing and well researched complex of the three buildings alongside the churches on Ajdovski gradec form a whole that includes a representational, living and commercial building (Knific 1994, 212, 215 and 216; Ciglenečki 1995, 175-177). The numerous objects found in the small living quarters are carefully selected, high quality and very diverse as regards their functionality -all of which indicates that this was the living quarters of an important individual (Petru 1979, 731). The remaining researched living quarters in the settlement show an equal distribution of activities necessary in the everyday supply of the inhabitants; it is clear that they were used for living quarters as well as for commercial use (Vičič 1983, 71-76; Knific 1994, 215-217). The double church on Vranje was merely big enough for a few churchgoers, and these most likely lived within the accompanying settlement complex. At special occasions it might have served also the inhabitants from the surroundings. The lack of small objects in the representative heated room confirms that it was not v sklopu naselbine. Hiše so v celoti raziskane in večidel objavljene (Petru 1979; Knific 1979; 1986; 1994, 212, 214-216; Vičič 1983, 71; Mirnik - Prezelj 1984). Arheološko ovrednotenje kaže, da predstavlja v podrobnostih znan in zelo izpoveden sklop treh stavb ob cerkvah na Ajdovskem gradcu zaključeno celoto, ki obsega reprezentančen, bivalni in gospodarski del (Knific 1994, 212, 215 in 216; Ciglenečki 1995, 175-177). V majhnih bivalnih prostorih najdeni številni predmeti so izbrani, funkcionalno različni, a kakovostni, ter kažejo na bivališče pomembne osebe v naselbini (Petru 1979, 731). Tudi ostale doslej raziskane bivalne stavbe v tej naselbini kažejo enakomerno porazdelitev dejavnosti za vsakodnevno oskrbo prebivalcev v posameznih hišah; služile so torej tako bivanju kot tudi gospodarski uporabi (Vičič 1983, 71-76; Knific 1994, 215-217). Dvojna cerkev na Vranju je zadoščala komaj peščici vernikov, ki so najverjetneje prebivali kar v pripadajočem naselbinskem sklopu, ob posebnih priložnostih pa je morda služila tudi prebivalcem iz okolice. Pomanjkanje drobnih predmetov v reprezentančnem ogrevanem prostoru potrjuje, da ni bil namenjen množici ljudi, ampak ožjemu krogu pomembnih oseb. Sorodno strukturiranost kažejo bivalni prostori blizu zgornje dvojne cerkve na Sv. Hemi (Glaser 1991, 69-71). Zato bi glede na bližino obeh zgradb (loči ju le razdalja 3 m) smeli podobno kot na Ajdovskem gradcu videti v njih stavbno kombinacijo bivalnega in reprezentančnega značaja (Ciglenečki 1995, 177-178). Tudi osnovna razporeditev prostorov spominja na omenjeni bivalni kompleks na Vranju, kjer so dimenzije ogrevanega, ločenega prostora povsem v razmerju z manjšo površino naselbine in manjšimi cerkvami. Glede na dimenzije posameznih zgradb, arhitekturno zasnovo, izpostavljeno mesto v naselbini in bližino cerkva smo torej upravičeni na Kučarju, Vranju in Sv. Hemi domnevati posebno vlogo centralno ogrevanih prostorov znotraj bivalne arhitekture, namenjene kleru. Na Vranju, kjer je bila arhitektura zaradi omejenega prostora najbolj funkcionalno zasnovana, je mogoče videti, da ima stavba z ogrevanim prostorom dejansko najbolj izpostavljen položaj in leži najbližje obema cerkvama. Kakšna natančno je bila funkcija sorazmerno velikega ogrevanega prostora, doslej ni bilo mogoče določiti, vendar pomanjkanje drobnih najdb, predvsem keramike, ne govori za bivališče, velikost in centralno ogrevanje pa izpostavljata njegov pomen. Kot episkopij označen skupek zgradb in zidov na Lavantu je preslabo znan, da bi ga lahko primerjali s prej omenjenimi. Obstoj dvojne cerkve, kot tudi novejše odkritje zgodnjekrščanske cerkve pod cerkvijo sv. Ulrika, velikost celotne naselbine in bližina v dolini ležečega škofovskega mesta Aguntum govorijo v prid domnevi o škofovskem sedežu v zadnjem obdobju antike. Dodatno to tezo potrjuje opustitev mesta na ravnini in sočasen nastanek zgradb v odlično naravno zavarovani naselbini intended for larger numbers of people, but for a small circle of important individuals. The living quarters next to the upper double church on Hemmaberg show a similar structure (Glaser 1991, 69-71). Due to the proximity of the two buildings (they are merely 3 metres apart) we could assume a combination of living quarters and a representative building, similar to Ajdovski gradec (Ciglenečki 1995, 177-178). The basic layout of the rooms is reminiscent of the aforementioned living quarters on Vranje, where the dimensions of the heated, separated room, is in relation to the smaller settlement area and the smaller churches. Taking into consideration the dimensions of the individual buildings, the architectural base, the exposed location within the settlement and the proximity to the churches (on Kučar, Vranje and Hemmaberg) we can assume that the centrally heated rooms had a special role within the living quarters of the clergy. On Vranje, where the architecture was the most functionally planned (due to the lack of space), it is possible to see that the building with the heated room has the most exposed position and that it is the closest building to the two churches. The function of the relatively large heated room has so far not been ascertained; however the lack of small finds, especially pottery, makes us believe that it was not used as living quarters, while the size and central heating indicate its importance. On Levant, the buildings and walls commonly denoted as the episcopal residence have not been researched enough to be compared with the previously mentioned ones. The existence of the double church combined with the newer discovery of another early Christian church (under the church of St. Ulrich), the size of the settlement and the proximity of the episcopal town of Aguntum (in the valley) speak in favour of the assumption that this was the episcopal seat during the last Antique period. This hypothesis is additionally confirmed by the abandonment of the towns in the plains and the simultaneous emergence of buildings in an excellently naturally protected settlement on the hill (Alzinger 1977, 406-410; Glaser 1997, 143-146; Tschurtschenthaler 2003). As traces of two churches and a defensive wall were found it seems that the poorly researched post of Grazerkogel was also an ecclesiastical centre (Egger 1916, 106-109; Ciglenečki 1987a, 61; Glaser 1997, 121). Taking into account the exposed position on the low hill rising from the middle of the valley (and consequentially the poor natural protection) we can assume a similar fate as experienced by the post on Kučar: the church centre was relocated to a safer area. Especially important for ascertaining the functions of the double churches in fortified hilltop settlements is the example of the episcopal centre in Saben (Sabiona), mentioned in written sources. A poorly preserved double church (which has not been published in detail) has been discovered during the systematic research carried na hribu (Alzinger 1977, 406-410; Glaser 1997, 143-146; Tschurtschenthaler 2003). Zdi se, da med obravnavana cerkvena središča sodi tudi slabo raziskana postojanka Grazerkogel, kjer so odkrili sledove dveh cerkva in obrambnega zidu (Egger 1916, 106-109; Ciglenečki 1987a, 61; Glaser 1997, 121). Glede na močno izpostavljenost na nizkem hribu sredi doline in posledično slabo naravno zaščito bi smeli domnevati podobno usodo kot pri postojanki na Kučarju: cerkveni središči sta bili preseljeni na varnejši območji. Za ugotavljanje funkcij dvojnih cerkva v utrjenih višinskih naselbinah je posebno pomemben primer v virih izpričanega škofijskega središča v Sabnu (Sabiona). pri sistematičnih raziskavah so na vrhu višinske naselbine na skalnem osamelcu pod cerkvijo sv. Križa našli slabo ohranjeno dvojno cerkev, ki pa ni objavljena v podrobnostih. V preliminarnih objavah so jo izkopavalci različno datirali: po prvi različici sodi v 7. st. (Bierbrauer 1998, 210-214), po drugi pa že v 5.-6. st. (Nothdurfter 2003, 311- 314). Obstoj sorodno koncipiranih bivalnih prostorov ob dvojnih cerkvah opozarja na njihov posebni pomen za razumevanje celote in narekuje njihovo vključitev v poskuse interpretiranja sakralnih sklopov na vrhu naselbin. Kratek pregled dosedanjih interpretacij teh kompleksov kaže zelo pestro podobo. T. Ulbert je v objavi cerkva z Ajdovskega gradca leta 1975 sklenil svojo interpretacijo z mislijo, da namembnosti še ni mogoče zadovoljivo pojasniti. Zaradi majhnega grobišča in razmeroma velikih prostorov v cerkvah je zato med drugim omenil tudi možnost pribežališča za škofa in nekakšne "pražupnije" na tem področju (Ulbert 1975, 67-68). p. petru je leta 1979 glede na odkritje uteži in amfor v cerkvam najbližji stavbi sklepal na prisotnost škofa, in sicer zaradi kompetenc, ki so jih škofje imeli po odloku cesarja (Petru 1979, 731). Pri prvi interpretaciji ostalin s Kučarja je S. Ciglenečki zaradi značaja raziskanih bivalnih prostorov in izjemnosti dvojnih cerkva na Slovenskem in Avstrijskem tam domneval sedež škofa (Ciglenečki 1986, 146). Glede na časovno ujemanje opustitve bližnjega mesta Neviodunum (ob koncu 4. ali začetku 5. st.) z izgradnjo kučarskega kompleksa (nastanek okoli leta 400) se je zdel verjeten premik cerkvenih oblasti iz Nevioduna na Kučar, v Belo krajino, oddaljeno od pomembnejših prometnic. Vendar je to tezo danes mogoče upoštevati le pogojno: previdnost narekuje predvsem odkritje velike, ob času objave članka povsem neznane poznoantične naselbine v Črnomlju, v kateri smemo pričakovati tudi večje cerkve iz 5. in 6. st. (Mason 1998; 2006). Ob raziskavi Sv. Heme se je problematiki funkcije zgodnjekrščanskih središč v višinskih naselbinah z arheološkega vidika posvetil F. Glaser (Glaser 1991, 74-85). Tako je po podrobni analizi arhitekture in okrasja v cerkvah na Sv. Hemi prišel do sklepa, da je bilo to romarsko središče. Svojo analizo je osredotočil predvsem out at the top of the hilltop settlement on the isolated rock hill under the church of St. Cross. The preliminary publications provide various dates: in the first version it was dated into the 7th century (Bierbrauer 1998, 210214), while in the second version it was believed to have originated in the 5th or 6th century (Nothdurfter 2003, 311-314). The presence of similar living quarters (next to double churches) draws attention to their importance in the understanding of the whole and dictates that they should be included into the attempts of interpreting ecclesiastical complexes on top of settlements. A quick overview of the interpretations carried out so far shows an extremely diversified image. In the 1975 publication discussing the churches on Ajdovski gradec T. Ulbert concluded his interpretation with the thought that their role could not be satisfactory explained at the time. The small burial site and the relatively large areas inside the churches made him consider that this could be a refuge for the bishop or an early parish (Ulbert 1975, 67-68). Based on the discovery of the weights and amphorae in the building closest to the churches, P. Petru (1979) assumed the presence of a bishop, which could be possible as the emperor's decree gave bishops certain competencies (Petru 1979, 731). In the first interpretation of the remains from Kučar S. Ciglenečki assumed that this was the episcopal see. He assumed this due to the character of the researched living quarters and the scarcity of double churches in Slovenia and Austria (Ciglenečki 1986, 146). Taking into account that the nearby town of Neviodunum was abandoned at roughly the same time (at the end of the 4th or beginning of the 5th century) as the complex on Kučar was constructed (approximately 400 AD) it seemed likely that the church authorities moved far from the important roads, from Neviodunum to Kučar (into the Bela krajina region). However, today this hypothesis can be accepted only conditionally: caution is dictated by the discovery of the large Late Antique settlement in Črnomelj, which was unknown at the time the article was published, and in which we can expect large churches dating to the 5th or 6th century (Mason 1998; 2006). F. Glaser addressed the function of Early Christian churches in hilltop settlements during his archaeological research of Hemmaberg (Glaser 1991, 74-85). Following a detailed analysis of the architecture and church decorations at Hemmaberg, he reached the conclusion that this was a pilgrimage centre. His analysis focused on the treatment of the church architecture, however as a large part of the settlement has not been researched, it is impossible to pass a final judgement as regards its character. Glaser was of the opinion that Kučar and Ajdovski gradec were also pilgrimage centres (Glaser 1997, 77, 81). In his text on double churches in the territory covered by present day Slovenia R. Bratož introduced na obravnavo tamkajšnje cerkvene arhitekture, vendar pa zaradi neraziskanosti velikega dela naselbine danes ni mogoče podati dokončne sodbe o njenem značaju. Romarsko namembnost je pripisal tudi Kučarju in Ajdovskemu gradcu (Glaser 1997, 77, 81). R. Bratož je v svoji obravnavi dvojnih cerkva na sedanjem slovenskem območju v razpravo poleg arheoloških virov pritegnil tudi pisne vire (Bratož 1996). Ugotovil je, da se dvojne cerkve ne omenjajo v nobenem viru z vzhodnoalpskega, severnojadranskega ali zahod-nobalkanskega območja. V primeru Ajdovskega gradca je prišel do sklepa, da si moramo na njem predstavljati močnejše regionalno cerkveno središče, ki ga je občasno obiskoval škof. Prav tako si zaradi skromnosti cerkva ni mogel predstavljati večjega prihoda romarjev. Opozoril pa je na možno prisotnost menihov. Glede obstoja škofijskih sedežev omenja večkratne prepovedi ustanavljanja škofijskih sedežev v manjših krajih, iz česar posredno sklepa, da je do tega vendarle prihajalo. V luči vira, ki v salonski škofiji kljub prepovedi omenja ustanovitev treh novih škofij, ne izključuje možnosti, da so novonastali škofijski sedeži obstajali tudi na območju vzhodnih Alp. Ob zgoraj navedenih poskusih razrešitve problema namembnosti višinskih naselbin z dvojnimi cerkvami se je temu mogoče približati tudi z druge strani. Glede na dinamično naselbinsko situacijo na vzhodnoalpskem območju, kjer so bila mesta v veliki meri opuščena, se namreč zastavlja vprašanje, kje so v času druge polovice 5. st., v 6. st. in morda delno celo v 7. st. rezidirali škofje nekaterih vzhodnoalpskih mest, o katerih poročajo viri, a so bila po sredini 5. st. opuščena (Ciglenečki 1987b, 265-271; Bratož 1991, 26-27). Če vendarle skušamo odgovoriti na to vprašanje ob doslej ugotovljenih zanesljivih cerkvenih zgradbah iz tega časa, se kot potencialni škofovski sedeži zarisujejo prav obravnavana poznoan-tična višinska najdišča, na katerih so bile ob dvojnih cerkvah odkrite tudi pomembnejše bivalne zgradbe z reprezentančnimi prostori. Na območju mestnih teritorijev zgoraj obravnavanih postojank doslej niso bile odkrite naselbine, v katerih bi cerkvena in cerkvam pripadajoča bivalna arhitektura imela večji pomen. Nasprotno, v največ primerih gre za manjše, slabše okrašene stavbe, ob katerih ni zaslediti večjih bivalnih objektov (pregledno pri Ciglenečki 1987a, 135-139; Glaser 1997, 70-160). Premiki cerkvenih središč (hkrati s premikom prebivalstva z ogroženih ozemelj) so znani tudi v drugih delih vzhodnoalpskega območja, čeprav niso vedno sočasni. Najbolj znan primer takšnega premika - sicer časovno najpoznejši - je nedvomno prenos sedeža pa-triarhata iz ogrožene Akvileje v bližnji Gradež, na otok v težko dostopnem lagunskem predelu (Cuscito 2001). Podeželske dvojne cerkve, ki so nastajale postopno, a kažejo v določenih fazah zelo sorodno podobo, je mogoče najti v precejšnjem številu tudi na območju province Dalmacije. Nastajale so v ravninskem svetu written sources and combined them with archaeological ones (Bratož 1996). He ascertained that double churches were not mentioned in any source originating from the east Alpine, north Adriatic or west Balkans area. In the case of Ajdovski gradec he reached the conclusion that this was a strong regional ecclesiastical centre that was occasionally visited by the bishop. Due to the modest churches he could not imagine greater numbers of pilgrims. However, he did mention the possibility of the presence of monks. As regards the episcopal see he mentioned the multiple bans on establishing episcopal sees in smaller settlements, which lead him to believe that this occasionally took place. Taking into account the source which mentions that three new dioceses were established in the Diocese of Solona regardless of the ban, he does not exclude the possibility that newly formed episcopal sees existed also in the east Alpine area. Alongside the above stated attempts to solve the issue as regards hilltop settlements with double churches this can also be approached from the other side. Taking into account the dynamic settlement situation in the east Alpine area, in which numerous towns were abandoned, we face the question as to where did the bishops of certain east Alpine towns reside in the second half of the 5th, throughout the 6th and partially even in the 7th century. The sources report their presence, however these towns were abandoned after the mid 5th century (Ciglenečki 1987b, 265-271; Bratož 1991, 26-27). If we try to answer this question with the use of the so far ascertained reliable church buildings from the time, the potential for episcopal sees seem to be in the discussed Late Antique hilltop sites, at which double churches and representational living quarters were discovered. So far not a single settlement within the town territories of the above mentioned posts has been discovered that would reveal important churches or clergy living quarters. On the contrary, in most cases we are dealing with small, modestly decorated buildings alongside which no large living quarters could be found (for an overview see Ciglenečki 1987a, 135-139; Glaser 1997, 70-160). The moves of church centres (together with the move of the inhabitants from endangered territories) are also known in other parts of the east Alpine area, however, not at the same time. The best known example of such a move - also the last - was the move of the patriarchal seat from the endangered Aquileia to the nearby Grado, onto an island in the hard to access lagoon (Cuscito 2001). Large numbers of the gradually emerging provincial double churches that show an extremely similar image (in certain phases) can be found in the province of Dalmatia. These churches were built in the plains alongside unfortified, so far still un-researched smaller settlements: they were ecclesiastical centres that belonged to small provincial territories. As an example of a well-researched double church we could expose Sl. 4.14: Srima pri Šibeniku. Dvojna cerkev (Gunjača 2005, sl. 18). Fig. 4.14: Srima near Šibenik. Double church (Gunjača 2005, Fig. 18). poleg neutrjenih, doslej še ne podrobneje raziskanih manjših naselij: v njih je mogoče videti prav tako sakralna središča manjših podeželskih teritorijev. Kot primer dobro raziskanih dvojnih cerkva bi smeli izpostaviti že dolgo znano najdišče Srima Prižba s številnimi grobnicami in pokopi (sl. 4.14), ki so z različnimi dogradnjami postopno prerasle v kompleksen sklop (Gunjača 2005). Precejšnjo podobnost s skupkom cerkva na Tonovcovem gradu kažejo tudi cerkveni kompleksi v Ži-tomislicih (sl. 4.15), Polačah (sl. 4.16), Jabuki in Zenici (sl. 4.17; Cambi 2005, 60, 62, 63). Edina dvojna cerkev v utrjenem podeželskem kompleksu je bila odkrita v zadnji fazi močno utrjene poznorimske vile v Mogorjelu (Suic 1976, 240). V luči teh primerjav je danes cerkveni sklop na To-novcovem gradu težko dokončno opredeliti, kajti pripadajoči bivalni in eventualni reprezentančni prostori, ki se nakazujejo v bližini, še niso raziskani. Cerkvene zgradbe, kot smo videli pri analizi cerkvenega kompleksa, dopuščajo možnosti različnih interpretacij in namembnosti. the centre Srima prizba with the numerous graves and tombs which has, together with the various additional structures (Fig. 4.14), gradually transformed into a complex building (Gunjaca 2005). The church complexes in Zitomislici (Fig. 4.15), Polace (Fig. 4.16), Jabuka and Zenica (Fig. 4.17) also show similarities with the churches at Tonovcov grad (Cambi 2005, 60, 62, 63). The only double church in the fortified provincial complex belonged to the last phase of the strongly fortified Late Antique villa in Mogorjelo (Suic 1976, 240). In the light of these comparisons it is hard to precisely define the church complex on Tonovcov grad, for the accompanying living quarters and the possible representation rooms (that can be seen in the vicinity) have not been researched yet. As we have seen in the analysis of the church complex, the church buildings allow for the possibility of various interpretations and uses. Thus, it makes sense to wait until at least the buildings in the vicinity have been researched before we attempt to de- Sl. 4.15: žitomislici. Dvojna cerkev s pripadajočo bivalno zgradbo (Andjelic 1978, sl. 5). Fig. 4.15: žitomislici. Double church with accompanying living quarters (Andjelic 1978, Fig. 5). fine the functions of individual buildings. The multiple churches, the high quality construction and the various liturgical functions accompanied by the fact that there are no similar Early Christian complexes far around, lead to the conclusion that this was an important regional ecclesiastical centre. When the preliminary research results from Tonovcov grad were published, S. Tavano assumed that this settlement functioned as a refuge for the bishop from the nearby town of Forum Iulii (Tavano 1995, 6; 2000, 253). In the presentation of the researched churches Ciglenečki interpreted the church complex in the middle of the settlement as a regional centre that was visited by the bishop on special occasions, and he did not exclude the possibility that a bishop of a strongly endangered area could find refuge here (Ciglenečki 1997c, 24). The presence of the bishop can neither be confirmed nor denied: archaeological sources are in this aspect too modest and written records are non-existent. Sl. 4.16: Polače. Dvojna cerkev (Chevalier 1995, 108). Fig. 4.16: Polače. Double church (Chevalier 1995, 108). Sl. 4.17: Zenica. Dvojna cerkev (Basler 1972, sl. 137). Fig. 4.17: Zenica. Double cburcb (Basler 1972, Fig. 137). Zato se zdi smiselno počakati na raziskave vsaj še bližnjih objektov, da bi zanesljiveje opredelili funkcije posameznih zgradb. Številnost cerkva, kakovostna gradnja in različne liturgične funkcije pa ob dejstvu, da daleč naokoli ni podobnega zgodnjekrščanskega sklopa, dovoljujejo sklep o pomembnem cerkvenem regionalnem središču. ob objavi preliminarnih rezultatov raziskav Tonovcovega gradu je S. Tavano izrazil domnevo, da gre na tej naselbini za pribežališče škofa iz bližnjega mesta Forum lulii (Tavano 1995, 6; 2000, 253). Pri predstavitvi raziskanih cerkva je ciglenečki interpretiral cerkveni sklop sredi naselbine kot regionalno središče, ki ga je ob posebnih priložnostih obiskal škof, ni pa izključil tudi možnosti, da je tu našel zavetje škof s katerega močneje ogroženega območja (Ciglenečki 1997c, 24). Prisotnosti škofa danes ni mogoče potrditi niti zanikati: arheološki viri so v tem oziru preskromni, pisni manjkajo. 4.3 astronomska orientacija cerkva na tonovcovem gradu 4.3 astronomical orientation of churches at tonovcov grad Saša ČAVAL, Ivan ŠPRAJC 4.3.1 UVOD Namen tega prispevka je opozoriti na morebiten astronomski pomen orientacij zgodnjekrščanskih cerkva na Tonovcovem gradu, zato nekaj uvodnih besed o vsebini in potencialih arheoastronomskih preučevanj ne bo odveč. Arheoastronomijo, ki se je kot posebna arheološka disciplina oblikovala šele v zadnjih desetletjih, zanimajo vsi kulturni pojavi, povezani z opazovanjem neba, ne le same oblike eksaktnega astronomskega znanja, ki so sicer predmet zgodovine astronomije, temveč tudi družbena vloga teh spoznanj in mnogi drugi elementi kulture, od gospodarstva do religije. Arheoastronomija z raziskovanjem teh pojavov in razlaganjem odvisnosti kozmoloških in drugih koncepcij, povezanih z astronomskim znanjem, od naravnega in kulturnega konteksta pomembno prispeva k razumevanju posameznih preteklih družb, splošnih procesov kulturne evolucije in drugih širše zanimivih vprašanj, ki si jih zastavlja arheološka veda (Aveni 2003, 149 ss; Šprajc 1991, 3 s; 2001, 385 s). Zametki arheoastronomije segajo v čas, ko je bila tudi arheologija šele v povojih. Že v 18. stoletju so posamezniki, predvsem na prazgodovinskih megalitskih najdiščih v Angliji, opažali zanimivosti, ki naj bi bile vezane zlasti na gibanje Sonca. Velik korak naprej je pomenila knjiga The Dawn of Astronomy (1894), delo britanskega astronoma Normana Lockyera, ki je zagovarjal hipotezo o gradnji egiptovskih piramid in templjev po nekaterih astronomskih načelih. Do teh podatkov je prišel na podlagi terenskega dela v Egiptu, kjer je izmeril orientacijo templjev in njihovo usmerjenost primerjal z gibanjem Sonca in zvezd (Michell 1989). Naslednji mejnik je bila knjiga astronoma Geralda Hawkinsa Stonehenge decoded, ki je izšla leta 1965 in je botrovala nastanku vede, sprva imenovane astroarheologija. Hawkinsovo delo, v katerem je znamenito prazgodovinsko najdišče v Angliji razlagal kot astronomski observatorij, je vzbudilo burne polemike, ki so pripeljale do kritičnega pregleda dotedanjega dela in do izoblikovanja metodologije za "novo vedo", za katero se je poslej uveljavilo ime arheoastronomija (Aveni 1981, 3 ss; Šprajc 1991, 3 s; 2001; Ruggles 1999). 4.3.1 INTRODUCTION This paper calls attention to the possible astronomical significance of orientations of the Early Christian churches at Tonovcov grad, thus a few words concerning the contents and the potential of archaeoastronomical studies will not go amiss. Archaeoastronomy, which developed as a special archaeological subdiscipline during the past few decades, is concerned with all cultural phenomena related to the observation of the sky, not only with the exact astronomical knowledge, which is the subject of the history of astronomy, but also with the latter's social role, as well as with a number of other cultural manifestations, from economy to religion. By attempting to solve these issues and to explain the dependence of the cosmological and other astronomically-derived concepts on the natural and cultural context, arhaeoastronomy offers an important contribution to the understanding of past societies, of the processes of general cultural evolution, and of other broadly interesting questions posed by archaeology (Aveni 2003, 149 ff.; Sprajc 1991, 3 f.; 2001, 385 f.). The origins of arhaeoastronomy can be traced back to the time when archaeology itself was only starting off. Some peculiarities possibly linked to the movement of the Sun began to be noticed, especially on prehistoric megalithic sites in England, already in the 18th century. The book The Dawn of Astronomy (1894) written by the British astronomer Norman Lockyer represented a great step forwards, for he advocated the hypothesis that certain astronomical principles were involved in the construction of the Egyptian pyramids and temples. This hypothesis was developed on the basis of his fieldwork in Egypt, where he measured the orientations of temples and compared them with the movements of the Sun and stars (Michell 1989). Another important turning point was astronomer Gerald Hawkins' book Stonehenge decoded, published in 1965 and originating a new scientific field, at first called astroarchaeology. Hawkins' work, in which he described the famous prehistoric site in England as an astronomical observatory, raised vivid polemics that led to a critical overview of the work performed until then Glede na to, da so eden od najbolj tipičnih virov za arheoastronomska preučevanja orientacije oziroma usmeritve, ki jih opažamo tako v arhitekturi kot v drugih elementih arheološkega zapisa, velja opozoriti, da lahko med pionirje arheoastronomskih raziskav uvrstimo tudi našega fizičnega antropologa Boža Škerlja, ki je že v petdesetih letih prejšnjega stoletja obravnaval orientacijo grobov na nekaterih srednjeveških grobiščih pri nas (Škerlj 1952). V razpravi je razglabljal o pravilih orientacije grobov in skušal najti dokaze o morebitnih obdobjih, v katerih so ljudje umirali pogosteje. Danes se arheoastronomske raziskave, katerih razvoj je v zadnjih desetletjih močno napredoval, osre-dotočajo na najrazličnejše kulture in obdobja. Pomemben predmet preučevanj je tudi orientacija krščanskih cerkva. Po starih pisnih virih (Origen, Klement Aleksan-drijski, Tertulijan) naj bi zgodnjekrščanske skupnosti sledile antični tradiciji in štele usmerjenost proti vzhodu za sveto (Firneis, Köberl 1989, 430 s). Že Vitruvij je v svojem delu De Architectura v 1. stoletju pr. n. št. predlagal, da naj bi bili templji usmerjeni tako, da božanstvo gleda na zahod, častilci oziroma verniki pa naj bi bili obrnjeni proti njegovemu kipu, torej proti vzhodu; tako naj bi bilo videti, da vzhajajoče Sonce in tudi samo božanstvo bdita nad njimi (McCluskey 1993, 112 s). "Templji bi morali biti usmerjeni proti tistemu delu neba, od koder je zemlja osvetljena," piše v zgodnjem 2. stoletju Hyginus Gromaticus v svojem delu Constitutio limitum (McCluskey 2007, 335 s). Prvi ekumenski koncil, ki je bil leta 325 v Niceji, je med drugim določil usmerjenost cerkva in s tem oltarja proti vzhodu ter izdal pravilo, da naj bo tudi duhovnik pri molitvi v cerkvi obrnjen proti vzhodu (Firneis, Köberl 1989, 430 s). Poleg naštetih pišejo o usmeritvi cerkva proti vzhodu tudi poznejši viri (McCluskey 2007, 335 s). Pri orientaciji zgodnjekrščanskih cerkva opažamo dve varianti: ali je bil oltar postavljen v vzhodni glavni apsidi cerkve, ki ima vhod na zahodni strani, ali pa je vhod v cerkev postavljen na vzhodni strani tako, da je prvi jutranji žarek posijal na oltar, ki je bil postavljen na zahodnem delu cerkve. Le malo cerkva je usmerjeno proti pravemu oziroma kardinalnemu vzhodu, o vzrokih za odmike od te smeri pa so mnenja različna. Pisni viri navajajo samo splošno pravilo o usmeritvi proti vzhodu, zato pri pojasnjevanju posameznih orientacij niso v pomoč, avtorji arheoastronomskih raziskav v Evropi pa menijo, da so bile cerkve orientirane: - v smeri pravega oziroma kardinalnega vzhoda ali proti ekvinokcijskemu vzhajališču Sonca; - v smeri vzhajajočega Sonca na god zavetnika cerkve; - v smeri Sončevega vzhoda ob enem ali drugem solsticiju (naravna mejnika v astronomskem letu, hkrati pa tudi dva izmed največjih krščanskih praznikov: božič in šentjanževo); and to the formation of methodology for a 'new discipline, which became known as archaeoastronomy (Aveni 1981, 3 ff.; Šprajc 1991, 3 f.; 2001; Ruggles 1999). Taking into account the fact that the orientations embedded in architecture and other elements of archaeological record represent one of the most typical data sources for archaeoastronomical studies, it should be recalled that one of the pioneers of archaeoastronomi-cal research was the Slovenian physical anthropologist Božo Škerlj, who as early as the 1950s studied grave orientations in certain Medieval cementeries in Slovenia (Škerlj 1952). By studying the orientation trends in burials, he tried to find evidence about the seasons in which mortality may have been particularly high. The archaeoastronomical research, having made a significant progress during the recent decades, is nowadays focused on various cultures and periods. Its relevant subjects include the study of orientations of Christian churches. According to old written sources (Origen, Clement of Alexandria, Tertullian), the early Christian communities followed the antique traditions in considering the orientation towards the east as sacred (Firneis, Koberl 1989, 430 f.). Vitruvius, in his work De Architectura (1st century BC), suggested that the temples should be oriented in such a way that the deity's statue facing west can be adored by the worshippers turned to the east, making it appear that the rising Sun as well as the deity are watching over them (McCluskey 1993, 112 f.). Hyginus Gromaticus in his early 2nd century work Constitutio limitum suggests that the temples should be oriented to 'that part of heaven from which the earth is illuminated' (McCluskey 2007, 335 f.). The first Ecumenical council in Nicaea, in AD 325, established that the churches should be oriented with the altar to the east, and decreed that the priest, while praying, should also look toward the east (Firneis, Koberl 1989, 430 f.). Apart from those already mentioned, later sources also write about church orientation (McCluskey 2007, 335 f.). Two orientation variants can be noticed among Early Christian churches: the altar was either placed in the eastern main apse and the entrance to the church was on the west side, or else, the entrance was placed on the eastern side so that the first morning rays of the Sun shone upon the altar located in the western part of the church. Only few churches were oriented to the true or cardinal east, and there is no consensus about the reasons for deviations from this direction. Since the available written sources only mention the general rule prescribing the orientation to the east, they are of no use in the attempts to explain individual orientations. According to different researchers in Europe, the churches were oriented: - towards the true or cardinal east or to the point where the Sun rose at the equinoxes; - to sunrises on the church's patron day; - to sunrises at one of the two solstices (two naturally significant moments of the astronomical year, and - v smeri vzhajajočega Sonca na dan, pomemben za ustanovitelja meniškega redu (posebej v cistercijanskem in frančiškanskem redu); - v smeri vzhajajočega Sonca na kak datum, pomemben ob izgradnji posamezne cerkve; - v smeri vzhoda glede na magnetni sever (Guszik 1978; Cave 1950; Firneis, Koberl 1989; Barlai 1989; Incerti 2001; Iwaniszewski 1998, 189 ss; 1999, 109 s; Benson 1956; Ministr 1997; Gattin, Pejakovič 1982; McCluskey 1993, 109 ss; 2007; Abrahamsen 1992). Kot kaže, je bila usmerjenost cerkva pomembna le do konca romanike, v nekaterih primerih pa tudi v zgodnji gotiki (Firneis, Koberl 1989, 431). pomembno je dodati, da navedena mnenja povečini ne temeljijo na sistematičnih raziskavah, ki bi zajemale večje število cerkva na posameznem območju, ali na dovolj natančnih podatkih, ki bi omogočali zanesljivo analizo orientacij in interpretacijo njihovega morebitnega astronomskega pomena. Med redkimi izjemami je članek McCluskeyja (2007), ki obravnava 130 srednjeveških cerkva v srednji in vzhodni Angliji, v sklepni fazi pa je tudi raziskava, ki temelji na natančno izmerjeni orientaciji 174 romanskih cerkva v Sloveniji (Čaval 5. a.). Rezultati v obeh primerih kažejo, da so bile cerkve orientirane proti Sončevim pozicijam na horizontu na nekatere pomembne cerkvene praznike, povezava med godovi zavetnikov cerkva in datumi, ki ustrezajo orientaciji, pa ni pogosta. 4.3.2 USMERJENOST CERKVA also two important Catholic holidays: Christmas and St. John's); - to the sunrise on a day important to the founders of a monastic order (especially in the Cistercian and Franciscan orders); - to the sunrise on an important date linked to the construction of a specific church; - to the magnetic east (Guszik 1978; Cave 1950; Firneis, Koberl 1989; Barlai 1989; Incerti 2001; Iwaniszewski 1998, 189 ff.; 1999, 109 f.; Benson 1956; Ministr 1997; Gattin, Pejakovic 1982; McCluskey 1993, 109 ff.; 2007; Abrahamsen 1992). Whatever the true reasons were, it appears that the orientation of the churches was important only until the end of the Romanesque period, or in certain cases also in the Early Gothic period (Firneis, Koberl 1989, 431). It is important to add that the aforementioned opinions in most cases do not derive from systematic studies based on a large number of churches in a particular area, nor on precise alignment data that would allow reliable analyses of orientations and interpretations of their possible astronomical meaning. A rare exception is Mc-Cluskey's article (2007) discussing 130 Medieval churches in central and eastern England, while a research based on precisely measured orientations of 174 Romanesque churches in Slovenia is currently in its final phase (Caval s.a.). The results of both studies show that the churches were oriented to the Sun's positions on the horizon on important church holidays; however, the relations between the dates corresponding to the orientations of the churches and their patron days are relatively uncommon. Na Tonovcovem gradu so bile zgrajene tri cerkve. Severna in osrednja cerkev (brez narteksov) sta bili postavljeni hkrati, konec 5. ali v začetku 6. stoletja. Sredi 6. stoletja so prenovili notranjost obeh, dodali katedre in naredili estrihe. Takrat ali nekaj pozneje je bila na skrajnem južnem delu skalnega platoja prizidana še ena cerkev in narteks k prvima (Ciglenečki 2006; 2008, 511 ss). Izmere orientacij so bile opravljene pri vseh treh cerkvah. Pri meritvah s teodolitom in astronomsko referenco (Soncem) je bila uporabljena ustaljena tehnika, ki omogoča izmero in izračun pravih (astronomskih) azimutov1 (Thom 1971, 120 ss; Aveni 2001, 120 ss; Ruggles 1999, 164 ss; Šprajc 1991, 45 s). Že iz tlorisa najdišča (sl. 1.7) je jasno vidno, da imata severni cerkvi enako usmeritev, orientacija južne pa je drugačna. podatki o obeh orien- 1 Azimut je kot, merjen v horizontalni ravnini od severa proti desni ali - gledano od zgoraj - v smeri urnih kazalcev, in ima vrednosti od 0° do 360°. Magnetni azimuti, dobljeni s kompasom, se nanašajo na magnetni sever, ki se ne ujema z geografskim oz. astronomskim, zato niso uporabni za ugotavljanje morebitnega astronomskega pomena orientacije, razen če jih popravimo za vrednost lokalne magnetne dekli-nacije (Šprajc 1999, 16 s). 4.3.2 ORIENTATION OF THE CHURCHES Three churches were erected at Tonovcov grad. The north and main churches were built, both without a narthex, in the same period, i.e. in the late 5th or in the early 6th century. In the mid 6th century both interiors were renovated, cathedrae and a mortar floor were added. At the same time (or a little later) another church was built on the southernmost part of the rocky plateau, and narthex was added to the two existing churches (Ciglenečki 2006; 2008, 511 ff.). The orientations of the three churches were measured with a theodolite and astronomical reference (the Sun), employing the established technique that enables the measurement and calculation of true (astronomical) azimuths1 (Thom 1971, 120 ff.; Aveni 2001, 120 ff.; Rug- ' The azimuth is the angle measured in the horizontal plane from the north towards the right or - as seen from above - clockwise, and has values ranging between 0° and 360°. The magnetic azimuths, obtained with a compass, refer to the magnetic north, which does not coincide with the geographical or astronomical one, and thus cannot serve for ascertaining the eventual astronomical meaning of orienta- Tab. 4.1: Podatki o orientaciji cerkva na Tonovcovem gradu. Tab. 4.1: Church orientation data at Tonovcov grad. Stavba / Building A h 8 julijanski datum / Julian date severna in osrednja cerkev / north and main churches 135°27' 315°27' 2°33' 9°25' -27°36' 37°06' - južna cerkev / south church 121°39' 301°39' 6°21' 12°32' -16°23' 30°40' 2. februar / February 2nd, 5. november / November 5th tacijah so navedeni v tabeli 4.1. Azimuti, ki opredeljujejo obe orientaciji, predstavljajo srednje vrednosti azimutov, merjenih vzdolž več zanesljivih stavbnih linij, ohranjenih in situ. Ker je vsaka usmeritev lahko funkcionalna v dveh smereh, sta za vse navedena po dva azimuta, ki se razlikujeta za 180°; vrednosti se nanašata na različne astronomske pojave - nekateri so vidni na vzhodnem horizontu in drugi na zahodnem - in ustrezajo jima različne višine horizonta (h) in deklinacije2 (5), zapisane v naslednjih dveh stolpcih. Tako kot azimuti so bile tudi ustrezne višine horizonta izmerjene na terenu, deklinacije pa izračunane z uporabo formul sferne trigonometrije (Hawkins 1968, 50 ss; Šprajc 1991, 48 s; Thom 1971, 120 ss; Ruggles 1999); geografska širina, potrebna pri teh izračunih, je bila izmerjena s prenosno GPS-napravo (N46°15'27"; E13°34'50"). Ker je deklinacija, ki ustreza orientaciji južne cerkve proti vzhodu, znotraj obsega Sončevih deklinacij, sta v tem primeru navedena tudi ustrezna julijanska datuma Sončevega vzhoda, ki veljata za 6. stoletje n. št., torej za čas, ko je bila cerkev zgrajena.3 Glede na to, da so merjeni zidovi sorazmerno kratki in ne povsem ravni, in upoštevaje odstopanja med azimuti posameznih linij, merjenih na vsaki stavbi, je mogoče oceniti, da je možna napaka v deklinacijah približno ±1°, kar pomeni, da je v datumih, ki ustrezata orientaciji južne cerkve proti vzhodu, možna napaka ± 3 dni. 2 Da bi lahko identificirali astronomski pojav, na katerega se utegne nanašati določena orientacija, je treba izračunati deklinacijo, ki ustreza izmerjenemu azimutu in višini horizonta ter geografski širini opazovalca. Vsako točko v prostoru (nebesno telo, točko na horizontu ...) je mogoče projicirati na imaginarno nebesno sfero, deklinacija točke na tej sferi pa izraža njeno kotno oddaljenost od nebesnega ekvatorja, ki si ga lahko zamišljamo kot projekcijo Zemljinega ekvatorja na nebesno sfero. Deklinacije se merijo pravokotno na nebesni ekvator proti severu in jugu in imajo vrednosti od 0° do ±90°. S pomočjo deklinacije točke na horizontu lahko ugotovimo, katerim nebesnim telesom ustreza (ali je ustrezalo v preteklosti) to vzhajališče ali zahajališče. 3 Deklinacija Sonca se čez leto nenehno spreminja in ob solsticijih doseže ekstremne vrednosti, ki so bile v 6. stoletju n. št. okoli ±23°37'. Za določitev datumov, ki ustrezajo posameznim Sončevim deklinacijam v preteklosti, je treba uporabiti primerne efemeride, npr. Ephemeris Generator, ki ga je izdelal Jet Propulsion Laboratory, NASA (ZDA), in je dostopen na spletu (http://ssd.jpl.nasa.gov/cgi-bin/eph). gles 1999, 164 ff.; Sprajc 1991, 45 f.). As clearly visible on the site map (Fig. 1.7), the north and main churches share the same orientation, while the south church is aligned differently. The alignment data for the two churches are given in Table 4.1. The azimuths defining the two orientations represent the mean values of the azimuths measured along a number of reliable and approximately parallel building lines preserved in situ. As any orientation can function in two directions, two azimuths (with a difference of 180°) are given for each orientation. The two values may be related to different astronomical phenomena, some visible on the eastern and others on the western horizon; the corresponding horizon altitudes (h) and declinations2 (S) appear in the next two columns. Like the azimuths, the horizon altitudes were measured in the field, while the declinations were calculated employing formulae of spherical trigonometry (Hawkins 1968, 50 ff.; Sprajc 1991, 48 f.; Thom 1971, 120 ff.; Ruggles 1999); the geographic latitude, necessary for these calculations, was determined with a portable GPS device (N46° 15'27"; E13°34'50"). As the declination corresponding to the eastward orientation of the southern church is within the range of solar declinations, the Julian dates of sunrise in that direction are also provided; they are valid for the 6th century, when the church was built.3 Since the meas- tions, except if we correct them for the value of the local magnetic declination (Sprajc 1999, 16 f.). 2 In order to identify the astronomical event that an orientation might refer to, the declination corresponding to the measured azimuth, the horizon altitude and the observer's geographical latitude must be calculated. Any point in space (a celestial body, a spot on the horizon, etc.) can be projected onto an imaginary celestial sphere, and the declination of the point on this sphere expresses its angular distance from the celestial equator, which can be conceived as a projection of the Earth's equator onto the celestial sphere. Declinations are measured perpendicularly to the celestial equator towards the north and south and have values ranging from 0° to ±90°. Having the declination of a point on the horizon, it is possible to ascertain which celestial bodies rise or set (or used to do it in the past) at that point. 3 The declination of the Sun changes throughout the year, reaching its extreme values at the solstices; in the 6'h century AD these were approximately ±23°37'. For determining the dates corresponding to particular solar declinations in the past, appropriate ephemerides have to be used, e.g. the Ephemeris Generator elaborated by the Jet Propulsion Labora- Obe deklinaciji, ki ustrezata orientaciji severne in osrednje cerkve, sta onstran vrednosti, ki jih dosega Sonce. Velja pa omeniti, da orientacija proti vzhodu približno ustreza vzhajališču Lune ob največjem južnem ekstremu. Deklinacija Lune se v enem mesecu (obhodu okoli Zemlje) močno spreminja in doseže skrajno pozitivno in negativno vrednost, ekstremne deklinacije pa se spreminjajo skozi daljši čas: največji ekstremi se ponavljajo vsakih 18,6 leta, v enakih intervalih pa je mogoče opazovati tudi najmanjše ekstreme (o velikosti in periodičnosti lunarnih ekstremov glej razlage v: Thom 1971, 16 ss; Morrison 1980, S65 ss; Šprajc 1991, 27 s). Orientacija obeh severnih cerkva na Tonovcovem gradu ustreza na vzhodnem horizontu Lunini deklinaciji -26°48'4, medtem ko je bila okoli leta 500 n. št., ko naj bi bili obe cerkvi zgrajeni, najmanjša Lunina deklinacija -28°47'. Čeprav je razlika skoraj 2°, namerne povezave te orientacije z Luno ni mogoče povsem izključiti: ker ne v Sloveniji ne drugje niso bile opravljene sistematične raziskave usmerjenosti zgodnjekrščanskih cerkva, trenutno ne vemo niti, kateri astronomski motivi in v kolikšni meri so narekovali orientacijo, zato tudi ne poznamo stopnje preciznosti, s katero so bile želene smeri - če so bile namerne - inkorporirane v arhitektonske elemente.5 Alineacije proti velikim in malim lunarnim ekstremom so pogoste na prazgodovinskih megalitskih najdiščih zahodne Evrope (Thom 1971; Morrison 1980; Ruggles 1999; Belmonte Aviles, Hoskin 2002), preliminarni rezultati študije, ki je v teku, pa kažejo, da je z lunarnimi ekstremi mogoče povezati tudi orientacijo več romanskih cerkva v Sloveniji (Čaval 5. a.). Če bodo statistične analize razkrile, da naključnost teh ujemanj ni verjetna, bi lahko domnevali, da so bile nekatere cerkve namerno usmerjene proti lunarnim ekstremom: če upoštevamo velik pomen Lune v krščanstvu (za določanje datuma velike noči), takšna hipoteza ni povsem neverjetna. Spomnimo se, da zgodnji pisni viri, četudi orientacijo cerkva povezujejo s Soncem, metaforično podobo Kristusa (v. supra), o natančnih pravilih usmerjanja ne povedo ničesar. Druga, morda verjetnejša razlaga pa je, da so bile nekatere cerkve zgrajene na ruševinah predkrščanskih svetišč: če so ta bila tudi na slovenskem prostoru pogosto usmerjena proti lunarnim ekstremom, kot je dokazano za velik del prazgodovinske Evrope, je mogoče domnevati, da so cerkve to orientacijo ohranile, ker so bili pri gradnji uporabljeni posamezni zidovi ali 4 Vrednost je nekoliko drugačna od deklinacije v tabeli 4.1, ker je pri izračunu Luninih deklinacij treba upoštevati paralakso (Hawkins 1968, 50 ss; Šprajc 1991, 48 s). V primeru ostalih nebesnih teles, ki so od Zemlje veliko bolj oddaljena kot Luna, je paralaksa zanemarljiva. 5 Velja dodati, da so zidovi severnih dveh cerkva na Tonovcovem gradu dokaj neravni in le približno vzporedni med seboj (sl. 1.7), zaradi česar ni mogoče natančno ugotoviti, katera smer je bila graditeljem najpomembnejša. Srednji azimut v tabeli 1 temelji predvsem na meritvah vzdolž vzhodnih segmentov zidov, ki so med seboj bolj vzporedni kot ostali. ured walls are relatively short and not entirely straight, and considering discrepancies between the azimuths of particular lines measured at each building, the possible error in declinations has been estimated to be of about ±1°, which means that the error in the dates corresponding to the eastward orientation of the southern church is of up to ±3 days. Both declinations corresponding to the orientation of the north and main churches are beyond the values reached by the Sun. It is worth noting, however, that the orientation towards the east roughly fits the rising point of the Moon at its major southern standstill. Within a period of one month, the declination of the Moon varies greatly, reaching its extreme positive and negative values, but the extreme declinations exhibit long-term variations: the major extremes or standstills occur every 18.6 years, and the same intervals separate the minor standstills (for explanations of the extents and periodicity of lunar standstills, see: Thom 1971, 16 ff.; Morrison 1980, S65 ff.; Sprajc 1991, 27 f.). On the eastern horizon the orientation of the northern two churches at Tonovcov grad corresponds to the Moon's declination of -26°48'4, while at the time around AD 500, when the two churches were built, the minimum lunar declination was -28°47'. Even though the difference is almost 2°, an intentional link between this orientation and the Moon cannot be completely discarded: in view of the lack of systematic studies of orientations of Early Christian churches, both in Slovenia and elsewhere, we do not know which astronomical phenomena, and to what extent, may have motivated the orientations, and thus we also have no reliable information about the precision with which the desired directions - if intentional - were incorporated into the architecture.5 Alignments to major and minor lunar standstills are common in prehistoric megalithic sites throughout Western Europe (Thom 1971; Morrison 1980; Ruggles 1999; Belmonte Avilés, Hoskin 2002), and the preliminary results of the study currently underway show that the orientations of several Romanesque period churches in Slovenia may also be linked to lunar extremes (Caval s.a.). If statistical analyses reveal that these correspondences are not fortuitous, we could assume that tory, NASA (USA) and available online (http://ssd.jpl.nasa. gov/cgi-bin/eph). 4 The value is slightly different from the declination given in Table 4.1, because the parallax has to be taken into account when calculating lunar declinations (Hawkins 1968, 50 ff.; Sprajc 1991, 48 f.). In the case of celestial bodies that are much more distant from the Earth than the Moon, the parallax is negligible. 5 It should be added that the walls of the northern two churches at Tonovcov grad are rather irregular and only roughly parallel to each other (Fig. 1.7), which makes it impossible to precisely ascertain the direction that was relevant to the builders. The mean azimuth in Table 1 is based largely on readings taken along the eastern wall segments, which are more parallel to each other than others. temelji prejšnjih stavb. Čeprav na Tonovcovem gradu ostanki predkrščanskih stavb niso bili najdeni, je poselitev tega prostora v prazgodovini nedvomno izpričana (glej pogl. 2.2.1 in Tonovcov grad. Najdbe, pogl. 6). Deklinacija, ki ustreza orientaciji južne cerkve proti zahodu, se ne zdi pomenljiva, vzhodno deklinacijo pa je mogoče povezati s Sončevimi vzhodi okoli 2. februarja in 5. novembra v julijanskem koledarju 6. stoletja. Pomen 5. novembra v cerkvenem koledarju ni očiten; zaradi ocenjene napake bi se lahko ta orientacija sicer nanašala na 2. ali celo 1. november, toda praznika vseh svetih in vernih duš sta se začela slaviti na ta dva datuma šele v 9. in 10. stoletju (Ginzel 1914, 192 s; Ušeničnik 1945, 268 ss; Metford 1991, 115 ss). Zato je veliko bolj verjetno, da se orientacija nanaša na 2. februar, liturgični praznik, imenovan svečnica ali Jezusovo darovanje, katerega praznovanje sega v pozno antiko. Na svečnico se blagoslavljajo sveče in izvajajo obhodi z njimi, kar je prazniku tudi dalo ime. Cerkev časti ta dan zaradi očiščevanja Marije (Purificatio Beatae Mariae Virginis) in darovanja otroka Jezusa v templju, kar se je po starem judovskem zakonu opravljalo štirideseti oziroma osemdeseti dan po rojstvu otroka6 (Ušeničnik 1945, 200 s; Holweck 1908). V Jeruzalemu so praznovali svečnico že v prvi polovici 4. stoletja, a ne 2., temveč 14. februarja, torej štirideseti dan po 6. januarju, ki je takrat na Vzhodu veljal za dan Jezusovega rojstva. Ko je bil v zadnji četrtini 4. stoletja za Jezusov rojstni dan sprejet 25. december, se je tudi svečnica premaknila in se od takrat dalje praznuje 2. februarja (Ušeničnik 1945, 200 s; Čuk 1999, 56 s; Holweck 1908; Meier 2002, 90 s). Najzgodnejši vir o praznovanju svečnice je Itine-rarium Peregrinatio, delo nune Egerie (tudi Etheria, Echeria), ki se je med letoma 381 in 384 podala na romanje v Sveto deželo. Egeria poroča o procesiji in maši na 14. februar, ki sicer še nima imena,7 vendar lahko obred prepoznamo kot liturgično vsebino praznovanja svečnice (Bechtel 1910; Fraser 1994). Cedrenus, bizantinski zgodovinar iz 11. stoletja, pravi, da so v Antiohiji svečnico praznovali že leta 526, za Konstantinopel in celoten vzhodni imperij pa je praznik zaukazal cesar Justinijan leta 542 in se od takrat praznuje 2. februarja (Ušeničnik 1945, 200 s; Holweck 1908; Metford 1991, 75 s). Praznik je opisan v Gelazijevem 6 Po judovski postavi je morala vsaka mati, ki je rodila otroka, štirideseti oziroma osemdeseti dan po porodu iti v tempelj, tam darovati jagnje in se tako očistiti. Če je rodila sina, je bila po porodu nečista 7 dni in je morala počakati še 33 dni, če je rodila hčerko pa 14 in nato nadaljnjih 66 dni. V tem času se ni smela družiti z nikomer, zlasti pa ni smela v svetišče (Holweck 1908; Ovsec 1992, 92 s). 7 V zapisih, ki se nanašajo na svečnico, Egeria piše le o "... quadragesimae de epiphania (Fraser 1994). Gre za epifanijo oziroma praznik sv. treh kraljev, ki se je v Zahodni cerkvi že praznoval 6. januarja, medtem ko se je v večini 4. stoletja v Vzhodni cerkvi na ta dan praznovalo Jezusovo rojstvo (Martindale 1909). some churches were deliberately oriented to the lunar extremes: considering the importance of the Moon in Christianity (for determining the date of Easter), such a hypothesis is not totally unlikely. It should be recalled that the early written sources - even if they link the practice of church orientation with the Sun, i.e. the metaphorical image of Christ (v. supra) - are not explicit about the exact orientation rules. Another, possibly more likely explanation is that some churches were built on the ruins of pre-Christian temples: if the latter were, also in Slovenia, often oriented to lunar extremes (as has been proven for a number of sites in prehistoric Europe), we could assume that these churches incorporated parts of the walls and bases of previous buildings, thereby also preserving their orientations. Even though no remains of pre-Christian buildings have been found at Tonovcov grad, there is indisputable evidence of prehistoric occupation of the area (see chapter 2.2.1 and Tonovcov grad. Finds, chapter 6). While the declination corresponding to the westward orientation of the southern church does not seem significant, the eastern declination can be related to sunrises around February 2 and November 5 in the 6th century Julian calendar. The importance of November 5 is not obvious in the church calendar; due to the estimated error this orientation could also refer to November 1 or 2, but it was only in the 9 th and 10 th centuries that All Saints and All Souls days started to be celebrated on these two dates (Ginzel 1914, 192 f.; Usenicnik 1945, 268 ff.; Metford 1991, 115 ff.). It is thus much more likely that the orientation refers to February 2, a liturgical feast named Candlemas or Presentation of Jesus at the Temple, the celebration of which goes back to Late Antiquity. During Candlemas the candles were blessed and processions with candles took place, and this was what gave the holiday its name. The church celebrates this day as the Purification of the Blessed Virgin Mary (Purificatio Beatae Mariae Virginis) or the Presentation of Baby Jesus at the Temple. According to the old Jewish tradition this was performed on the fortieth or eightieth day after the child was born6 (Usenicnik 1945, 200 f.; Holweck 1908). In Jerusalem Candlemas was celebrated already in the first half of the 4th century; however, this did not take place on 2nd but rather on 14th February, i.e. on the fortieth day after 6th January, which was celebrated as Christ's birthday in the East of the time. In the last quarter of the 4th century, when December 25th was accepted as Christ's birthday, 6 According to Jewish tradition every mother who gave birth to a child had to go to the temple on the fortieth or eightieth day after the child was born to sacrifice a lamb, with which she became pure once more. If she gave birth to a son she was impure for 7 days after giving birth and had to wait for 33 days, and if she gave birth to a daughter she was impure for 14 days and had to wait next 66 days. During this period she was not allowed to socialise with anybody, and she was especially not allowed to enter a temple (Holweck 1908; Ovsec 1992, 92 f.). zakramentu iz 6. stoletja pod novim imenom očiščenje blažene device Marije (Purificatio B.V.M.). Na Zahodu je svečnica bolj Marijin praznik, saj poudarja spomin na darovanje prvorojenca Bogu, na Vzhodu pa Jezusov, ker poudarja njegovo spoznanje za Mesijo (Oražem 1989, 90 ss; Holweck 1908). Svečnica je praznik luči in naj bi prekrila poganska praznovanja. Cerkev je hote ali nehote upoštevala starodavna, a še aktualna izročila, močno tradicijo nekaterih praznikov ter hrepenenje človeka po svetlobi, še posebej v zimskem času, in je na ta dan uvedla blagoslov sveč in obhode z gorečimi svečami. Rimljani so februarja praznovali praznik luperkalij, ko so bile na ulicah množice ljudi z gorečimi plamenicami; v drugih sprevodih, amburbalijah, pa so prav tako s plamenica-mi plašili in podili duhove rajnih, ki jim je potekel čas obhoda na tem svetu (Ušeničnik 1945, 201 s; Ovsec 1992, 92 s; Kuret 1989, 524 ss). Luperkalije8 so bile zelo star, verjetno še predrimski praznik rodovitnosti in očiščevanja. Po slednjem nosi mesec februar tudi ime (februare = "očistiti"). Dan luperkalij je bil znan tudi kot dies februatus - očiščevalni dan (Green 1931, 62 ss). Božanstvo, kateremu naj bi bile luperkalije posvečene, ni znano. Po različnih virih naj bi pripadal praznik Favnu, Panu, Luperku, Liceju (Lycaeus), Inuu (Inuus), Silvanu, Bakhu, Junoni itd., vendar o tem ni soglasja. Nekateri celo menijo, da naj bi bile luperkalije sprva magičen obred, ki torej ni bil posvečen nobenemu božanstvu. S tem lahko tudi razložimo, zakaj se je ohranil tako dolgo v čas krščanstva (Green 1931, 64 ss; Wiseman 1995, 1 ss). Šele leta 494 je papežu Gelaziju uspelo ukiniti luperkalije9 in jih po nekaterih virih prekriti s takrat še quadragesima Epiphaniae, vendar ni jasno, v kakšni datumski povezavi naj bi bile luperkalije s svečnico: luperkalije so slavili dva dneva po februarskih idah, torej 15. februarja, kar je sicer blizu 14. februarju, vendar pa se svečnica oziroma Marijino očiščenje v Rimu ni nikoli praznovalo 14. februarja, saj je bil praznik premaknjen na 2. februar preden se je uveljavil v Rimu (Green 1931, 60 s). Drugače pa je z amburbalijami (lat. ambio: grem okoli; urbs: mesto). Rimska kultna procesija naj bi se od nekdaj izvajala v začetku februarja. Čeprav nekateri avtorji omenjajo procesijo, ki je šla skozi mesto s prižganimi plamenicami in naj bi se nanašala na slavljenje umrlih (Ušeničnik 1945, 201 s; Ovsec 1992, 92 s; Kuret 1989, 524 ss; Gueranger 1904, 533 s), drugi pa govorijo o amburbalijah kot o prazniku mestnega obzidja (De Coulanges 1873, 132 s; Kanekar 1992, 48 ss; Weber 2004, 16 s), se vsi strinjajo v datumu praznovanja. Bolj kot vprašanje izvora svečnice in njene povezave s predkrščanskimi obredi je v našem primeru pomemb- the Candlemas was moved to February 2 (Ušeničnik 1945, 200 f.; Čuk 1999, 56 f.; Holweck 1908; Meier 2002, 90 f.). The earliest records on the celebration of Candlemas can be found in Itinerarium Peregrinatio, a work by nun Egeria (also known as Aetheria), who set on a pilgrimage to the Holy Land between the years 381 and 384. Egeria reports on a procession and mass on February 14, which does not yet have a name,7 but the ritual can be recognised as a liturgical celebration of Candlemas (Bechtel 1910; Fraser 1994). Cedrenus, an 11th century Byzantine historian, affirms that Candlemas was celebrated in Antiochia as early as AD 526, while for Constantinople and the rest of the Eastern Empire the holiday was enforced by Emperor Justinian in 542, and since that time it has been celebrated on February 2 (Ušeničnik 1945, 200 f.; Holweck 1908; Metford 1991, 75 f.). In the 6th century Gelazi's sacrament the holiday is described under the new name Purification of the Blessed Virgin Mary (Purificatio B.V.M.). In the West, Candlemas is considered to be Mary's holiday, for it emphasises the memory of the presentation of the firstborn to God, while in the East it is rather dedicated to Christ, underscoring its nature as a Messiah (Oražem 1989, 90 ff.; Holweck 1908). Candlemas, as a celebration of light, was intended to suppress pagan celebrations. The church (deliberately or not) took into account the strong ancient - yet still practised - tradition of certain festivals and the human longing for light, especially during winter, by introducing the blessing of the candles and a procession with lit candles on that day. In February the Romans celebrated the feast of lupercalia, when the streets were crowded with people carrying burning torches. In other processions, amburbalia, the torches were used to frighten and scare away the spirits of the deceased whose time for wandering around this world has passed (Ušeničnik 1945, 201 f.; Ovsec 1992, 92 f.; Kuret 1989, 524 ff.). Lu-percalia 8 was a very old, possibly pre-Roman feast of fertility and purification. February obtained its name from the latter (februare = 'purify '). The day of luper-calia was also known as dies februatus - purification day (Green 1931, 62 ff.). The deity presiding lupercalia is not known. According to different sources, the feast was dedicated to Faunus, Pan, Lupercus, Lycaeus, Inuus, Silvan, Bakh, Junona, etc. As some authors say, lupercalia may have been originally a magical ritual and therefore 8 Ime praznika oziroma rituala naj bi izviralo iz imena jame (Luperkal), kjer naj bi volkulja hranila Romula in Rema (Wiseman 1995, 2 s). 9 Po drugih virih je bil to papež Feliks III. (Wiseman 1995, 17 s). 7 In the records related to Candlemas, Egeria only wrote about '...quadragesimae de epiphania^ '. (Fraser 1994). It is the Epiphany or the holiday of the Holy Three Kings, which was in the Western Church already celebrated on 6th January; however, throughout most of the 4'h century the Birth of Christ was celebrated by the Eastern Church on this day (Martindale 1909). 8 The name of this holiday or ritual supposedly originates from the cave (Lupercale) in which the she-wolf fed Romulus and Remus (Wiseman 1995, 2 f.). no dejstvo, da je praznovanje 2. februarja ekleziastično uzakonil cesar Justinijan leta 542 (v. supra). Njegova želja, da bi ponovno združil nekdanje rimsko cesarstvo, se je zdela zelo uresničljiva sredi 6. stoletja, ko je bizantinska vojska premagala Ostrogote in dve leti pozneje zavrnila tudi Franke. Justinijan, ki je bil velik privrženec krščanstva in je pogosto celo sam sodeloval v debatah o krščanski doktrini, naj bi močno promoviral Marijine praznike10, kar svečnica tudi je. Posebno pomenljivo je, da je bil v prezbiteriju osrednje cerkve na Tonovcovem gradu najden Justinijanov srebrnik, kovan med letoma 548 in 552, in da naj bi bila v tistem času ali malo pozneje zgrajena tudi južna cerkev (Ciglenečki 2008, 511 ss, 519 s): takšna datacija vsekakor podpira domnevo o namernosti orientacije proti Sončevemu vzhodu na praznik, katerega čaščenje je bilo pravkar uzakonjeno. 4.3.3 SKLEP Dejstvo, da se med astronomskimi pojavi, s katerimi je mogoče povezati cerkve na Tonovcovem gradu, zdijo pomenljivi samo tisti, ki ustrezajo orientaciji proti vzhodu, je v skladu z zgodnjimi pisnimi viri, ki poudarjajo pomen vzhoda pri usmerjanju cerkva (Firneis, Köberl 1989, 430; McCluskey 2007, 335). Kljub temu je treba poudariti, da so zaradi pomanjkanja komparativnih in kontekstualnih podatkov predložene interpretacije nujno hipotetične. Če želimo z zadostno mero zanesljivosti potrditi, da je imela neka arhitektonska usmeritev namerno izbrano astronomsko tarčo, potrebujemo statistično pomenljivo število primerljivih usmeritev, nanašajočih se na isto pozicijo (deklinacijo) na nebesni sferi in vključenih v koherenten vzorec arheoloških elementov (torej istega tipa in pripadajočih istemu kulturnemu kompleksu), ali neodvisne kontekstualne podatke, ki sugerirajo astronomski motiv za to orientacijo (ikonografijo, pisne vire), ali oboje. Po drugi strani pa je pomen orientacij z isto astronomsko referenco mogoče primerno razumeti samo, če najdemo razloge, zaradi katerih je utegnil biti postulirani astronomski fenomen zanimiv za družbo, ki je usmeritev materializirala (Aveni 2003; Iwaniszewski 1989; Ruggles 1999). V našem primeru lahko zatrdimo, da so lahko bili astronomski pojavi, na katere se nanašajo tukaj predložene interpretacije, pomenljivi za zgodnjekrščansko oz. poznoantično družbo. Posebej velja opozoriti na konkretne zgodovinske okoliščine, ki podpirajo hipotezo o namernosti orientacije južne cerkve na Tonovcovem gradu proti Sončevemu vzhodu na 2. februar julijanskega koledarja v 6. stoletju: če je bila cerkev zgrajena po letu 542 - in takšna datacija je po razpoložljivih kronoloških podatkih verjetna -, je mogoče v njeni orientaciji videti materialen dokaz pomena svečnice, katere čaščenje je bilo v bizantinskem imperiju, del katerega je zagotovo 10 Codex lustinianus 1,27,9 (543). not dedicated to any particular deity, which could explain why it survived so long into Christianity (Green 1931, 64 ff.; Wiseman 1995, 1 ff.). It was only in 494 that Pope Gelasius managed to abolish lupercalia9 and substitute this feast, according to some sources, with quadragesima Epiphaniae, but the relationship between the dates of lupercalia and Candlemas remains unclear: lupercalia was celebrated two days after the February ides, i.e. on February 15, which is close to February 14, but Candlemas or Purification of the Blessed Virgin Mary was never celebrated on February 14 in Rome, for the holiday, before it arrived there, had been moved to February 2 (Green 1931, 60 f.). The case is different with amburbalia (lat. ambio: going around; urbs: town). The Roman cult procession supposedly always took place at the beginning of February. Even though some authors mention a procession that went through town with lit torches and was supposedly linked to the celebration of the dead (Ušeničnik 1945, 201 f.; Ovsec 1992, 92 f.; Kuret 1989, 524 ff.; Guéranger 1904, 533 f.), while others describe amburbalia as a feast related to the town walls (De Coulanges 1873, 132 f.; Kanekar 1992, 48 ff.; Weber 2004, 16 f.), both agree as to the date of these celebrations. Rather than the origin of Candlemas and its relation to pre-Christian rituals, it is important, in our case, that the ecclesiastical celebration on February 2 was enforced by emperor Justinian in 542 (v. supra). His ambition to reunite the former Roman Empire seemed viable in the mid 6th century, when the Byzantine army defeated the Ostrogoths and two years later also managed to fend off the Franks. Justinian, who was a great follower of Christianity and often participated in debates on the Christian doctrine, is believed to have promoted Mary's holidays10, among which Candlemas also belongs; furthermore, this was the time when the area of Tonovcov grad was under Byzantine government. It seems particularly significant that a Justinian silver coin minted between AD 548 and 552 was found in the apse of the main church at Tonovcov grad, and that in this period, or slightly later, the southern church was built (Ciglenečki 2008, 511 ff., 519 f.): such dating renders a strong support to the hypothesis that this church was intentionally oriented to sunrise on the date decreed as a feast only a few years earlier. 4.3.3 CONCLUSION Among the astronomical phenomena that can be related with the churches at Tonovcov grad, only those corresponding to the eastward orientations seem significant. This is in accordance with the early written 9 According to other sources this was Pope FeJix III (Wiseman 1995, 17 f.). 10 Codex Iustinianus 1.27.9 (543). vsaj nekaj časa bil tudi Tonovcov grad, tistega leta uradno uvedeno. Ker pa nimamo eksplicitnejših kontekstualnih podatkov, bo mogoče veljavnost predloženih hipotez ovrednotiti samo, če bodo opravljene sistematične raziskave dovolj velikega števila orientacij zgodnjekrščanskih cerkva. Upati je, da bodo rezultati raziskave orientacije romanskih cerkva v Sloveniji (Čaval 5. a.) osvetlili problem, čeprav ostaja vprašanje, v kolikšni meri bo mogoče pravila usmerjanja cerkva v romaniki ekstrapolirati na zgodnjekrščansko obdobje. records, which emphasise the importance of the east for the church orientations (Firneis, Koberl 1989, 430; McCluskey 2007, 335). It should be stressed, however, that the interpretations given above are, due to the lack of comparative and contextual data, inevitably hypothetical. In order to suggest, with a reasonable degree of confidence, that an architectural orientation had a deliberately chosen astronomical target, we need a statistically significant number of comparable orientations referring to the same position (declination) on the celestial sphere and incorporated into a coherent sample of archaeological features (i.e. of the same type and belonging to the same cultural complex), or independent contextual data suggesting an astronomical motive underlying the orientation (iconography, written records), or both. On the other hand, the significance of orientations with the same astronomical referent can be properly understood only if we find reasons for which the postulated astronomical phenomenon may have been interesting to the society that produced the alignments (Aveni 2003; Iwaniszewski 1989; Ruggles 1999). In our case it can be affirmed that the astronomical phenomena involved in the interpretations proposed above may have been meaningful to the local Early Christian society. Particular attention should be drawn to specific historical circumstances that support the intentionality of the orientation of the southern church at Tonovcov grad to sunrises on February 2nd of the 6th century Julian calendar: if the church was built after AD 542 - which is likely in the light of the available chronological data - its orientation may represent a material evidence of the importance of Candlemas, which in that year was introduced as an official holiday in the Byzantine Empire, whose territorial control evidently included the area of Tonovcov grad, at least for a while. In the absence of more explicit contextual data, only a systematic research of a sufficiently large number of early Christian church orientations can shed light on the validity of the proposed hypothesis. Hopefully the issue can be clarified by the results of the ongoing study of orientations of the Romanesque churches in Slovenia (Caval s.a.), even though the question remains as to the extent to which the rules of church orientation in the Romanesque period can be extrapolated to the early Christian period. 5. vloga in pomen poznoanticne naselbine 5. the role and importance of the late antique settlement 5.1 Utrdba Tonovcov grad - pomemben člen pozno- 5.1 The fort at Tonovcov grad - an important part of rimske obrambe Italije the Late Roman defence system of Italy 5.2 Tonovcov grad v poznoantični poselitveni sliki 5.2 Tonovcov grad in the broader Late Antique vzhodnoalpskega in zahodnobalkanskega prostora settlement pattern of the East Alpine and Western Balkans 5.1 utrdba tonovcov grad -pomemben člen poznorimske obrambe italije 5.1 the fort at tonovcov grad -an important part of the late roman defence system of italy 5.1.1 UVOD Prvo pomembnejšo sklenjeno poselitev na Tonovcovem gradu je mogoče umestiti v drugo polovico 4. st. in v prva tri desetletja 5. st. Kot smo pokazali v prejšnjih poglavjih, pričajo o tem najdbe keramike, skromni arhitekturni ostanki ob stavbi 1 in preostanek stavbe 3 ter kovinski predmeti in novci, najdeni pretežno v mlajših poznoantičnih plasteh. Nastanek in vlogo poznorimske utrdbe na Tonov-covem gradu je mogoče razumeti le, če jo umestimo v sočasno poselitveno sliko širšega območja, skozi katero so vodile poti v Italijo z najbolj ogrožene vzhodne in severovzhodne strani, in hkrati v dinamične vojaško-po-litične razmere tedanjega časa. Utrdba je nastala namreč takrat, ko je bila Italija, kot smo poučeni iz vojaškega priročnika Notitia dignitatum, zavarovana v Alpah z varovalnim sistemom Tractus Italiae circa Alpes (Šašel 1956; 1971, 37-38; Christie 1991, 418-420). Njegov najbolj izpostavljeni del na jugovzhodnem območju Alp, dodatno utrjen s sistemom zapornih zidov, utrdb in stolpov, je v virih večkrat omenjen in ga najustrezneje označuje izraz Amijana Marcelina Claustra Alpium luliarum (Šašel, Petru 1971; Christie 1991, 413-420; Horvat 1999, 231-232). Poznorimsko utrdbo na Tonovcovem gradu je treba zato tako geografsko kot tudi funkcionalno umestiti v prostor med Emono in Tarsatiko na vzhodni ter Akvilejo in Forum lulii na zahodni strani, prostor, ki ga je Jaro Šašel v svojih analizah označil kot vojno krajino (Šašel 1971, 38). Gre za območje, na katerem je prav v pozno-rimskem obdobju prišlo do močnega porasta trajneje obljudenih postojank na naravno zavarovanih hribih, ki pa jim zaradi slabe raziskanosti in neznačilne arhitekture ni mogoče vedno zanesljivo ugotoviti natančnejše funkcije (Ciglenečki 1987a, 114-115; 1997a, 193-195). Analiza novcev kot tudi drobnih predmetov pri doslej opravljenih sistematičnih raziskavah na Hrušici, na Lanišču in na Martinj hribu, ki so bili del zapornega sistema, poudarja pomen utrjene krajine predvsem v drugi polovici 4. st. (Kos 1986, 195-207; Leben, Šubic 1990; Pflaum 2007). Vzporejanje spektra poznorimskega 5.1.1 INTRODUCTION The first important permanent settlement on Tonovcov grad can be dated into the second half of the 4th and the first three decades of the 5th century. This is confirmed by the pottery finds, the modest architectural remains alongside building 1 and the remains of building 3 as well as the metal objects and coins, most of which were found in the later Late Antique layers. The origin and role of the Late Roman settlement at Tonovcev grad can only be understood if we view it within the contemporary settlement pattern of the broader area and take into account the dynamic military and political conditions of the time. The broader area was criss-crossed by routes that led towards Italy from the most endangered east and north east direction. The fort emerged at a time when Italy was - as explained in the military manual Notitia dignitatum - protected by the defensive system Tractus Italiae circa Alpes (Šašel 1956; 1971, 37-38; Christie 1991, 418-420). Its most exposed part in the Southeastern Alps, was additionally fortified by a system of walls, forts and towers and was in written sources most often and most appropriately named Claustra Alpium luliarum, a name coined by Ammianus Marcellinus (Šašel, Petru 1971; Christie 1991, 413-420; Horvat 1999, 231-232). The Late Roman fort at Tonovcov grad had to be geographically and functionally placed between Emona and Tarsatica on the east and Aquileia and Forum Iulii on the west, an area marked as a military territory in the analysis of J. Šašel (1971, 38). In the Late Roman period this area experienced a high increase in inhabited posts on naturally protected hills, however due to the poorly researched and atypical architecture it is impossible to reliably ascertain their exact function (Ciglenečki 1987a, 114-115; 1997a, 193-195). The analysis of coins and small finds during the systematic research on Hrušica, Lanišče and Martinj hrib - all of which were a part of the Alpine defensive system - emphasises the importance of the fortified territory during the second half of the 4th century (Kos 1986, 195-207; Leben, Šubic 1990; Pflaum 2007). When gradiva s Tonovcovega gradu s tistim z že bolje znanih vojaških postojank pokaže, da gre za identičen nabor značilnih predmetov moške noše (glej Tonovcov grad. Najdbe, pogl. 2.1.3). O tem priča poleg zgoraj omenjenih najdišč tudi npr. gradivo z Gradišča pri Dunaju, Zbelovske gore, Gradca pri Prapretnem in Limberka (Ciglenečki 1994a, sl. 2, 5; t. 1, 3, 6, 10a; 2007, t. 1, 2). Izredno številne in bogate elemente moške noše, hkrati pa skromne ostanke arhitekture iz te faze na Tonovcovem gradu je mogoče razložiti le s prisotnostjo vojaške posadke. Glede na v literaturi večkrat izraženo dilemo o značaju moške oprave, ki jo tolmačijo kot vojaško opremo ali tudi kot del opreme civilnih uslužbencev, pomeni najdiščni kontekst na Tonovcovem gradu pomembno oporo tezi o njegovem vojaškem značaju (Pflaum 2002, 275-276). Lega naselbine na izrazito strateškem, a gospodarsko nepomembnem območju priča o poudarjeno vojaškem značaju tukajšnje moške oprave. Tudi za druge podobne postojanke lahko na hribih, kjer ni bilo značilnih trajneje poseljenih naselbin, se pa pojavlja koncentracija najdb moške oprave, domnevamo njihov pretežno vojaški značaj, saj bi bila civilna uporaba tako izpostavljenih točk v takšni množini sicer povsem nerazumljiva. Tudi uvoženo keramiko, ki je na Tonovcovem gradu številna, v zadnjem času raziskovalci vedno bolj povezujejo z organizirano oskrbo vojske - anono (Vidrih Perko, Župančič 2003, 463-467). 5.1.2 OMREŽJE RIMSKIH CEST IN POTI NA OBMOČJU JULIJSKIH ALP we compared the Late Roman material from Tonovcov grad and the material from better known military posts we observed that Tonovcov grad showed an identical scope of characteristic male attire objects (see Tonovcov grad. Finds, chapter 2.1.3). Apart from the previously mentioned sites this was also confirmed by the finds at the following sites (amongst others): Gradišče near Dunaj, Zbelovska gora, Gradec near Prapretno and Limberk (Ciglenečki 1994a, Figs. 2, 5; Pls. 1, 3, 6, 10a; 2007, Pls. 1, 2). At Tonovcov grad the plentiful and rich male attire and the modest architectural remains from this period can only be explained with the presence of a military garrison. Taking into account the commonly expressed dilemma as regards the character of the male attire, which is explained as a part of the military equipment or as a part of the costume used by civil servants, the finds at Tonovcov grad offer strong support to those who believe in its military character (Pflaum 2002, 275276). The position of the settlement on a strategically important, but economically completely irrelevant area leans in favour of attributing a military character to the discovered objects of male attire. Similarly, we can assume a predominantly military character also for other similar posts on hills with no typical long-term settlements, but which show a concentration of male attire finds, for it would be completely incomprehensible that such exposed points would be used in such quantities by civilians. Research also linked imported pottery - which has been found in large quantities also on Tonovcov grad - to the organised supply of the army - annona (Vidrih Perko, Župančič 2003, 463-467). za boljše razumevanje nastanka poznorimske utrdbe na Tonovcovem gradu in predvsem povečanega pomena za njim ležečega Čedada (Cividale) je treba opozoriti na razvejano mrežo komunikacij, ki so omogočale dostop v Italijo čez prelaze: pomembne so tako postale vse ceste in poti, ki so vodile severneje in južneje od dobro vzdrževane državne ceste čez Hrušico, posebej še potem, ko je bila ta v začetku 5. st. opuščena (prim. pri Ciglenečki 1985, 267-270; 1997a, 186, 188-189). Številni ostanki zapornih zidov v zahodni Sloveniji jasno kažejo, da so obstajale stranske poti, ki jih je bilo treba v določenem trenutku varovati ali zapreti, čeprav jih miljniki ne označujejo in niso bile vrisane v tedanje itinerarije (Šašel 1975, 87, 105). Za nobeno drugo območje nimamo tako prepričljivega dokaza o obstoju mreže cest in poti: če nam npr. za vzhodno slovenijo manjkajo -razen utrdb - vsi elementi za tovrstno rekonstrukcijo, pa jo tukaj dobro poznamo prav po zaslugi posameznih odsekov zapor. Ti potrjujejo obstoj sočasnih in največkrat tudi starejših poti in praviloma ležijo na trasi cest, ki v severozahodnem delu Slovenije še danes najzanesljiveje in v najkrajši možni črti vodijo z Gorenjskega v Posočje. Zaradi hribovitega in goratega terena so bile pomembne 5.1.2 THE ROMAN ROAD NETWORK IN THE JULIAN ALPS For a better understanding of the origin of the Late Roman fort on Tonovcov grad - and especially for a better understanding of the growing importance of Cividale in its background - we have to draw attention to the diversified communication network that enabled Italy to be entered across various mountain passes: thus, all roads and routes that led to the north or south of the well maintained state road across Hrušica became important, especially once this state road was abandoned in the beginning of the 5th century (see Ciglenečki 1985, 267-270; 1997a, 186, 188-189). The numerous remains of defensive walls in western Slovenia clearly show the presence of side roads that had to be protected or closed even though they were not marked by milestones and were not included into the itineraries at the time (Šašel 1975, 87, 105). We do not have such convincing evidence for the road network in any other area: for instance, in eastern Slovenia we lack all elements for such a reconstruction (except forts). Sl. 5.1: Poskus rekonstrukcije cest in poti iz poznorimskega obdobja v severozahodnem delu Slovenije ter sočasne postojanke in zapore. Fig. 5.1: An attempt to reconstruct the Late Roman roads, routes and posts in northwest Slovenia. iste trase, saj je možnost ugodnih prehodov čez hribe v kar največji meri narekovala potek komunikacij. Sistem zapor, ki je bil vzpostavljen za varovanje matične Italije na najbolj prehodnem območju vzhodnih Alp in njihovega predgorja, je bil največkrat uporabljen pravzaprav v državljanskih vojnah (Saria 1939, 145; Petru 1972, 344; 1978, 505-506). Zapore so ovirale dostop vojaškim enotam, ki so - že znatno manjše - uporabljale On the other hand, in western Slovenia the roads are clearly indicated by the individual preserved parts of the defensive walls. These remains confirm the existence of contemporary and most often also older routes and as a rule they run along the roads that are in northwest Slovenia even today the most applicable and the shortest routes between Gorenjska region and the Soča Valley. Due to the hilly and mountainous terrain the same tudi manj znane ali slabše grajene poti. Seveda pa je ta obrambni mehanizem deloval tudi v globino, na širokem območju pred zaporami in za njimi. Ceste skupaj z utrdbami na strateško pomembnih krajih omogočajo dovolj zanesljivo rekonstrukcijo tedanjih komunikacij. Dodaten element so tudi utrdbe v tesneh, na vhodu v ozke doline, kjer zaporni zid pravzaprav ni bil potreben: značilna primera sta prav Tonovcov grad in Puštal nad Trnjem (Ciglenečki 1987a, 89; Štukl 2004), zelo verjetna pa Gradišče nad Sotesko pri Koroški Beli (Valič 1997, 262-263) in Prvejk pri Dolenji Trebuši (neobjavljeno). V zvezi s Tonovcovim gradom je treba zato podrobneje predstaviti mrežo komunikacij, ki so v severozahodni Sloveniji in delno avstrijski Koroški in Furlaniji omogočale dostop čez alpski svet v furlansko ravnico in katerih pomembno stičišče je bilo prav v zgornjem delu Soške doline, na območju Kobarida. Na priloženem zemljevidu je poskus ponazoritve poznorimske mreže komunikacij na območju Julijskih Alp z vrisanimi zapornimi zidovi, sočasnimi utrdbami in nekaterimi drugimi slabše znanimi poznorimskimi najdišči (sl. 5.1). Komentar k tako zasnovanemu zemljevidu je potreben le v toliko, da pojasni značaj posameznih utrdb in predstavi argumente, ki so omogočili zaris posameznih tras. CESTA čez HRUŠICO Vse od avgustejskega obdobja pa do konca 4. st. je bila najpomembnejša prometna vez med severno Italijo in vzhodom cesta čez Hrušico, kar je razvidno iz pisnih virov kot tudi arheoloških raziskav (Leben, Šašel 1971, 93-96; Ulbert 1981). Njeno traso jasno označujejo na več mestih odkrito cestišče in pomembne obcestne postaje in utrdbe ob njem. To vlogo je cesta izgubila prav ob koncu 4. ali najkasneje na začetku 5. st., kot to dobro nakazujejo najdbe iz sistematično raziskane trdnjave Ad Pirum (Ulbert 1981 46-49; Pflaum 2004, 149-153; sl. 5.2). Tako v sicer številnem novčnem gradivu na Hrušici, v Lanišču in na Vrhniki ni Honorijevih emisij novcev tipa Gloria Romanorum (prim. novčni obtok pri Kos 1986, 195-207). Posredno opustitev ceste čez Hrušico potrjuje tudi stolp v dolini pod postojanko, ki naj bi bil uničen leta 394 (Ulbert 1981, 35; Petru 1980-1981, 133). Zaporni zidovi in utrdbe ob tej glavni povezavi in cestah, ki so se navezovale nanjo, so bili opuščeni najpozneje po začetku 5. st. Prav ti podatki zato dokazujejo opustitev te trase in posredno oživitev drugih vzporednih poti proti Italiji, kar odlično odseva v drobnem gradivu in predvsem najpoznejših rimskih novcih iz številnih utrjenih poznorimskih postojank (Kos 1986, 216). čeprav večina avtorjev, ki omenja selitev Lango-bardov proti Italiji, domneva njihov premik prav po cesti čez Hrušico, zgoraj omenjene raziskave to možnost routes were used, for the communications were always dictated by the possibility of an easy pass across the hills and mountains. The system of barriers that was established in order to protect Italy at its most transient area in the Eastern Alps was in fact most frequently used in civil wars (Saria 1939, 145; Petru 1972, 344; 1978, 505-506). The barriers prevented the - significantly smaller - military units that used lesser known or poorer quality routes from entering the Italian territories. Of course, this defensive mechanism also worked in-depth; it ran across a broad area in front of the blockade as well as behind it. The roads and the forts placed on strategic locations alongside them allow for a sufficiently reliable reconstruction of the communication routes at the time. Additional points in the reconstruction of the communication system are represented by forts built on entrances into narrow valleys, where the defensive wall was unnecessary: typical examples are Tonovcov grad and Puštal above Trnje (Ciglenečki 1987a, 89; Štukl 2004), and most likely also Gradišče above Soteska near Koroška Bela (Valič 1997, 262-263) and Prvejk near Dolenja Trebuša (unpublished). In relation to Tonovcov grad we need to present in greater detail the communication network that in northwest Slovenia, Friuli and partially in Austrian Carinthia enabled access across the Alps and into the Friuli plain, with its important crossroads in the Upper Soča Valley, close to Kobarid. The map is an attempt to show the Late Roman communication network in the area covered by the Julian Alps, including the defensive walls, forts and some other lesser known Late Roman sites (Fig. 5.1). At this map we only need commentary in order to explain the character of the individual forts and present the arguments that have enabled the various routes to be drawn in. THE ROAD ACROSS HRUŠICA Written sources and archaeological research show that between the Augustan period and the end of the 4th century the most important transport route between northern Italy and the east lead along the road across Hrušica (Leben, Šašel 1971, 93-96; Ulbert 1981). The route this road took is clearly marked by the various parts of the road that were discovered and by the important roadside posts and forts. The road lost its role at the end of the 4th century or at the latest at the beginning of the 5th century, which is clearly shown by the finds from the systematically researched fort Ad Pirum (Ulbert 1981, 46-49; Pflaum 2004, 149-153; Fig. 5.2). The otherwise vast coin collections from Hrušica, Lanišče and Vrhnika do not include Honorius' coins type Gloria Romanorum (see coin circulation in Kos 1986, zanikajo (prim. Kos 1955, 27; Werner 1962, 14; Menghin 1985, 97-98; Christie 1995, 63). Pomanjkanje vsakršnih najdb, ki bi kazale na uporabo te poti po začetku 5. st., jasno kaže, da so ti premiki potekali drugje (Ciglenečki 1985, 267-270). Sl. 5.2: Hrušica. Poznorimska utrdba (po Ulbert 1981, pril. 1). Fig. 5.2: Hrušica. Late Roman fort (after Ulbert 1981, appendix 1). 195-207). The abandonment of the road across Hrušica is also confirmed by the tower in the valley below the fort which is believed to have been destroyed in 394 (Ulbert 1981, 35; Petru 1980-1981, 133). The defensive walls and forts along the main road and the roads that ran off it were abandoned at the latest by the beginning of the 5th century. This provides additional proof that this road was abandoned and the parallel roads leading towards Italy were brought to life, which is excellently reflected in the small finds and especially in the Late Roman coins found in the numerous fortified Late Roman posts (Kos 1986, 216). Even though most authors who discuss the move of the Lombards towards Italy assume that they travelled along the road across Hrušica, the above mentioned research denies this possibility (see: Kos 1955, 27; Werner 1962, 14; Menghin 1985, 97-98; Christie 1995, 63). The lack of any finds that would indicate the use of this road after the beginning of the 5th century clearly shows that these movements had to take place elsewhere (Ciglenečki 1985, 267-270). THE ROAD CIVIDALE-PREDEL-CARINTHIA The road that needs to be studied if one wishes to understand the importance of the fort at Tonovcov grad led northwards from Cividale along the Nadiža (Natisone) river to Kobarid, along the Soča river, across the Predel/Predil pass and finally it joined the itinerary road Aquileia-Virunum in the valley of the river Fella (see Bosio 1991, 193-199; Panciera 1976, 154). So far it has not been confirmed by milestones, but traces of the road have been found at numerous locations (Rutar 1882, 9; Maggi, Žbona Trkman 2007; Magnani 2007). Its route is more or less defined by the terrain, for to a great extent it runs along the rivers of Nadiža (Natisone) and Soča, however the precise route is unknown - especially in the mountainous world. The road is indirectly confirmed also by the settlements along the Nadiža (Natisone) river and in the Kobarid area (Rupel 2005; Ciglenečki 1997b). S. Rutar wrongfully recognised the posts alongside the road Aquileia-Santicum mentioned in the itineraries from the time, and assumed that this is where the march of the Lombards into Italy took place (Rutar 1882, 9, 15; Rutar 1890, 206-207). The strongly emphasised role of the town Forum Iulii, especially during the time of the Lombards (Figs. 5.3, 5.4); and the settlements and forts mentioned in chapter 1.5 indicate that the road became increasingly important during Late Antiquity (Bonetto, Villa 2003; Vitri, Villa, Borzacconi 2006). Sl. 5.3: Cividale (Čedad). Potek poznoantičnega obzidja (po Bonetto, Villa 2003, 38). Fig. 5.3: Cividale. Late Antique city walls (after Bonetto, Villa 2003, 38). CESTA ČEDAD (CIVIDALE)-PREDeL-KOROŠKA THE ROAD ALONG THE UPPER SAVA VALLEY Cesta, ki je za razumevanje pomena utrdbe na tonovcovem gradu primarnega pomena, je vodila iz Čedada ob Nadiži in od Kobarida dalje proti severu ob Soči čez prelaz Predel ter se v dolini Bele združila z iti-nerarsko cesto Akvileja-Virunum (pregledno pri Bosio 1991, 193-199; Panciera 1976, 154). doslej ni potrjena z miljniki, na več krajih pa so naleteli na njene sledove (Rutar 1882, 9; Maggi, Žbona trkman 2007; Magnani 2007). Njen potek v grobem določa terenska situacija, saj v velikem delu teče po dolinah rek Nadiže in soče, natančnejši potek trase pa - posebej v gorskem svetu -ni znan. Posredno jo potrjujejo predvsem naselbine ob Nadiži in na kobariškem območju (Rupel 2005; Ciglene-čki 1997b). s. Rutar je ob njej zmotno prepoznal celo v The easiest way to bypass the Julian Alps Is to travel along the Upper Sava Valley, which was in the Late Roman period extensively used also by those who wanted to bypass the system of defensive walls and reach Italy. At tarvisio the road joined the itinerary road Aquileia-Santicum. The importance of the road along the Sava Valley was strengthened during the Late Antiquity, when it was protected by the fortress Carnium (Kranj; Fig. 5.5) which was the most important Late Antique post in the entire territory covered by present day Slovenia (Werner 1962; Stare 1980). Individual finds are beginning to indicate the possibility of a Late Roman settlement in Kranj (Vidrih Perko, Sagadin 2004, 217; Sagadin 2008, 176). to this we can add some newer, unpublished finds Sl. 5.4: Cividale (Čedad). Del raziskanega poznoantičnega obzidja (Bonetto, Villa 2003, sl. 1). Fig. 5.4: Cividale. A part of the researched Late Antique city wall (Bonetto, Villa 2003, Fig. 1). itinerarjih omenjene postaje ob cesti AkvHeja-Santicum in domneval pohod Langobardov v Italijo po njej (Rutar 1882, 9, 15; Rutar 1890, 206-207). Njen pomen se je okrepil v poznoantičnem času, o čemer priča predvsem močno povečana vloga mesta Forum lulii (posebej za časa Langobardov; sl. 5.3, 5.4), hkrati pa tudi v poglavju 1.5 omenjene naselbine in utrdbe (Bonetto, Villa 2003; Vitri, Villa, Borzacconi 2006). CESTA PO ZGORNJESAVSKI DOLINI Masiv Julijskih Alp je mogoče najlažje zaobiti po Zgornjesavski dolini, in to so dodobra izkoristili tisti, ki so se v poznorimski dobi želeli izogniti sistemu zapornih zidov in po ovinku priti v Italijo. Pri Trbižu se je cesta navezala na itinerarsko cesto Akvileja-Santicum. Pomen ceste po savski dolini se okrepi v pozni antiki, ko jo je varoval predvsem kastel Carnium (Kranj; sl. 5.5) kot najpomembnejša poznoantična postojanka na celotnem območju današnje Slovenije (Werner 1962; Stare 1980). Posamezne najdbe že nakazujejo možnost poznorimske poselitve na kranjskem pomolu (Vidrih Perko, Sagadin 2004, 217; Sagadin 2008, 176). Temu je mogoče dodati nekatere novejše, neobjavljene najdbe delov vojaške noše (fibul s čebulastimi zaključki, okov s krilci z izkopavanj leta 2007 v naselbini in grobišču; ustno M. Sagadin in J. Lux). Domnevati smemo, da je - sicer mnogo skromnejša - poznorimska utrdba stala na istem mestu kot poznejša utrjena naselbina iz 6. st. Tudi dalje proti severozahodu ležečo postojanko Gradišče nad Pivko pri Naklem je A. Valič označil kot poznorimsko postojanko: odkril je sledove močnega of military dress (crossbow fibulae and a strap end found during the 2007 excavations in the settlement and the cemetery; orally passed on by M. Sagadin and J. Lux). We can assume that the much more modest Late Roman settlement already stood on the same spot as the 6th century fortified settlement. The post Gradišče above Pivka pri Naklem, further towards the northwest, was classified by A. Valič as Late Roman (Valič 1968): he discovered traces of a strongly fortified tower on a slightly elevated hill in the middle of the plain, additionally protected by trenches and ditches (Fig. 5.6). A Late Roman phase could also be assumed at the non-researched Gradišče above Soteska near Koroška Bela, even though hardly anything is known about this site (Ciglenečki 1987a, 87; Valič 1997, 263). The ruins indicate a polygonal tower (additionally protected by a ditch) on an excellent strategic location at the entrance into the Upper Sava Valley, at the point where the slopes of the Karavanke mountains come the closest to the river. We only have modest data from the poorly researched Upper Sava Valley at our disposal, thus the non-researched Late Antique location in Rateče could also be important (Sagadin 1993, 140). THE ROUTE ACROSS BOHINJ The old prehistoric connections between Bohinj and the Soča Valley are clearly evident as well as confirmed through folk tradition (Gabrovec 1974, 304). The route became important in Late Roman times, which is clearly indicated by the large fort of Ajdovski gradec near Bohinjska Bistrica (Fig. 5.7), which with its position in Sl. 5.5: Kranj. Približen obseg poznoantične utrjene naselbine in poznoantični nekropoli (po Ciglenečki 2001, sl. 8). Fig. 5.5: Kranj. A rough outline of the Late Antique fortified settlement and Late Antique necropolis (afl;er Ciglenečki 2001, Fig. 8). stolpa na nekoliko vzvišeni vzpetini sredi ravnice (sl. 5.6), ki je bil dodatno zavarovan z jarki in okopi (Valič 1968). Poznorimsko fazo bi smeli domnevati tudi na še neraziskanem Gradišču nad Sotesko pri Koroški Beli, ki pa je slabo znano (Ciglenečki 1987a, 87; Valič 1997, 263). Ruševinski sledovi kažejo na stolp poligonalne oblike, ki je bil dodatno zavarovan z jarkom in strateško odlično umeščen pri vhodu v Zgornjesavsko dolino na kraju, kjer se pobočja Karavank povsem približajo Savi. Iz slabo raziskane Zgornjesavske doline so podatki skromni, zato utegne biti v tej zvezi pomembna še neraziskana poznoantična lokacija v Ratečah (Sagadin 1993, 140). SMER SKOZI BOHINJ O stari prazgodovinski povezavi med Bohinjem in Posočjem imamo številne materialne dokaze, posredno pa jo potrjuje ljudsko izročilo (Gabrovec 1974, 304). Pomembna je postala v poznorimskem času. Na to misel napeljuje predvsem obstoj velike utrdbe Ajdovski gradec pri Bohinjski Bistrici (sl. 5.7), ki s svojo izstopajočo lego sredi doline zapira in kontrolira pot (Gabrovec 1975, 164). Močno obrambno zidovje, novci in posamezne najdbe pričajo o njenem pomenu prav v poznorimski dobi (Kos 1988, 128-129; Mušič 1999, 370-376). Walter Schmid, ki je izkopaval na postojanki, je njen obstoj povezal s cesto in zaporami v dolini Bače (Gabrovec 1975, 164). Glede na relief in potek kasnejših cest se je ta cesta najverjetneje nadaljevala preko alpskih prelazov v dolino Bače, kjer je posredno potrjena z zapornim zidom (Petru 1971, 82-83). Omeniti moramo tudi skupek železnih in bronastih najdb ob njeni trasi v jami Ajdovska luknja pri Soteski v Bohinju, visoko nad dolino. Na podlagi novcev je najdišče datirano v drugo polovico 4. st. (Valič 1985). Možno je, da je šlo za občasno opazovalno postojanko in stražarsko mesto ali pa zgolj za pribežališče. O poseljenosti bližnjega višavja v poznorimski dobi priča kar nekaj visokogorskih najdišč (Ogrin 2006, 99-104). Cesto domneva tudi Joachim Werner, ko omenja, da je blejska nekropola na Pristavi ležala ob prometnici v Bohinj ter čez Bačo in Sočo v Cedad (Werner 1962, 14). Na svojem zemljevidu utrjenih mest in prometnic jo je vrisal pravzaprav kot edino cesto, ki je vodila z Gorenjske v Cedad (Werner 1962, 122, sl. 29). V. Šribar je pri obravnavi poznoantičnega grobišča v Podmelcu potrdil Wernerjevo smer, za kranjske Lan-gobarde pa domneval, da so šli v Italijo po Selški dolini (Šribar 1967, 386). \ r \ \ \ \ Križišče iskra \ - ") □ domnevno naselje / presumed settlement □ grobišče / cemetery D cerkev / church -- domnevna pot / presumed road „ CD the middle of the valley controls the road through the valley (Gabrovec 1975, 164). The strong defensive walls, coins and individual finds testify as to its importance in the Late Roman period (Kos 1988, 128-129; Mušič 1999, 370-376). Walter Schmid, who was conducting the excavations at the post, linked its existence to the road and walls in the Bača Valley (Gabrovec 1975, 164). Taking into account the relief and the path of the later roads this road most likely continued across the alpine passes into the Bača Valley, where it is indirectly confirmed by the defensive wall (Petru 1971, 82-83). We also have to mention the iron and bronze finds from the cave Ajdovska luknja near Soteska in Bohinj, high above the valley, which has been dated with the use of coins into the second half of the 4th century (Valič 1985). It is possible that this was an occasional observation and guard post or merely a refuge. There are quite a few high altitude sites that indicate that the nearby mountains were settled during the Late Roman period (Ogrin 2006, 99-104). A road was also assumed by Joachim Werner, when he mentioned that the necropolis Pristava in Bled was situated next to the road from Bohinj, across the Bača and the Soča rivers to Cividale (Werner 1962, 14). On his map of fortified towns and roads this was the only road that led from Gorenjska region to Cividale (Werner 1962, 122, Fig. 29). When discussing the Late Antique cemetery in Podmelec, V. Šribar confirmed Werner's road, however he assumed that the Lombards from Carnium travelled towards Italy along the Selška Sora Valley (Šribar 1967, 386). Sl. 5.6: Gradišče nad Pivko pri Naklem. Poznorimska utrdba (po Valič 1968, pril. 1). Fig. 5.6: Gradišče above Pivka pri Naklem. Late Roman fort (after Valič 1968, appendix 1). Sl. 5.7: Ajdovski gradec pri Bohinjski Bistrici. Poznorimska utrdba (po Valič 1997, sl. 7). Fig. 5.7: Ajdovski gradec near Bohinjska Bistrica. Late Roman fort (after Valič 1997, Fig. 7). SMER PO POLJANSKI DOLINI Po Poljanski dolini je potekala odlična naravna povezava z ugodnim prehodom na primorsko stran. Njen pomen potrjujeta v primorskem delu kar dve zapori. Prva je bila postavljena pri Novi Oselici, druga pa pri Zarakovcu (Šašel 1971, 82; Petru 1971, 82-84; Brank 1979). Potek poti v zahodnem delu ni jasen in dopušča več možnih tras. Simon Rutar omenja poleg drugih cest tudi to in njen približen potek (Rutar 1882, 12). v bližini te trase leži nekaj najdišč, ki pa so slabo znana in le s posameznimi elementi nakazujejo možnost obstoja naselbin ali utrdb v poznorimskem obdobju. takšno je npr. najdišče Gorenja Ravan, kjer naj bi bil najden amforast jermenski zaključek, ki je datiran v začetek 5. st. (Pflaum 2001, št. 57). Najdbe poznorimskih fibul s čebu-lastimi zaključki so znane iz bližine vasi Police (ustno D. Božič). v rimski naselbini Most na Soči so našli zakladne najdbe iz let 388 in 401, ki jasno kažejo na povezavo z velikimi spopadi v omenjenih letih (Kos 1986, 246; načrt rimskih ostankov pri Maggi, Žbona trkman 2007, 69). Strateško lego na sotočju Idrijce in trebuščice ima še povsem neraziskana postojanka Prvejk v Dolenji trebuši, od koder posamezna najdba poznoantične pasne spone dopušča možnost datacije na površini delno nakazanih stavb v to obdobje (neobjavljeno). SMER PO SELŠKI DOLINI zelo hipotetična trasa, pri kateri pa gre prav tako za staro naravno povezavo po dolini Selške Sore in nato čez prelaz Petrovo Brdo v dolino Bače, kjer se je združila s cesto skozi Bohinj. Njeno vlogo dobro nakazuje utrdba Puštal nad trnjem pri Škofji Loki, ki je bila zgrajena na strmem hribu tik pred vhodom v sotesko Selške Sore in s tem odlično varovala dostop v sotesko (sl. 5.8). Na utrdbi je bila ugotovljena močna poznorimska plast in podobno kot pri tonovcovem gradu tudi tu lahko domnevamo prvo intenzivnejšo poselitev v poznorimskem obdobju (Ciglenečki 1987a, 89; Štukl 2004; Šemrov 2004, 117, št. 53/1). 5.1.3 StRAtEŠKI POMEN KOBARIŠKEGA OBMOČJA v POZNORIMSKEM OBDOBJU v razvejeno mrežo poti in prehodov se v Posočju odlično vključuje utrdba na tonovcovem gradu, ki je izkoristila izjemno strateško lego in hkrati naravne zaščitne danosti terena. vendar pa to ni bila edina poznorimska postojanka v kobariški okolici. Območje, na katerem se je združilo kar nekaj pomembnih poti, po katerih je bilo mogoče skozi dolino Nadiže zaobiti glavne dostope v Italijo, je postalo v tem času izjemno občutljivo in pomembno. THE ROUTE ALONG THE POLJANSKA SORA VALLEY An excellent natural connection with a favourable pass to Primorska led along the Poljanska Sora Valley. Two walls confirm its importance in the Primorska region. The first was positioned at Nova Oselica, and the second at Zarakovec (Šašel 1971, 82; Petru 1971, 82-84; Brank 1979). In the west its layout is unclear and allows for a number of possible paths. Simon Rutar mentions this road and its approximate route (Rutar 1882, 12). A few lesser known sites lie in the vicinity of this route and they indicate the possibility of the existence of settlements or forts in the Late Roman period merely with a few elements. Such is for instance the site Gorenja Ravan, where an amphora-shaped strap end dated into the beginning of the 5th century was discovered (Pflaum 2001, No. 57). The finds of Late Roman crossbow fibulae are known from the nearby village of Police (orally passed on by D. Božič). In the Roman settlement Most na Soči coin hoards (dated to 388 and 401) were discovered and they clearly indicate a link with the extensive battles that took place in these two years (Kos 1986, 246; plan of Roman remains in Maggi, Žbona Trkman 2007, 69). The totally non-researched post Prvejk in Dolenja Trebuša had a strategic position at the confluence of the Idrijca and Trebuščica rivers. The individual Late Antique belt buckle find allows for the possibility of dating the buildings which are still non-researched but visible on the surface into this period (unpublished). THE ROUTE ALONG THE SELŠKA SORA VALLEY This is merely a hypothetical route, which represents an old natural connection along the Selška Sora Valley and across the pass Petrovo Brdo into the Bača Valley, where it joined the road through Bohinj. Its role is indicated by the fort Puštal above Trnje near Škofja Loka, which was built on the steep hill in front of the entrance into the gorge of the Selška Sora, thus forming an excellent defence of the gorge (Fig. 5.8). A thick Late Roman layer was discovered on the fort and similar as on Tonovcov grad we can assume the first more permanent settlement during the Late Roman period (Ciglenečki 1987a, 89; Štukl 2004; Šemrov 2004, 117, No. 53/1). 5.1.3 THE STRATEGIC IMPORTANCE OF THE KOBARID AREA IN THE LATE ROMAN PERIOD The fort at Tonovcov grad, which took advantage of the excellent strategic position and the natural defence possibilities offered by the terrain, fits excellently into the diversified network of roads and passes in the Sl. 5.8: Puštal nad Trnjem. Poznoantična utrdba (po Štukl 2004, sl. 1). Fig. 5.8: Puštal above Trnje. Late Antique fort (after Štukl 2004, Fig. 1). Dosedanje arheološke raziskave na območju Gradiča nad Kobaridom so potrdile, da je na starem prazgodovinskem gradišču in pobočjih pod njim tudi v rimskem obdobju nastala večja naselbina z ugotovljenim kultnim prostorom. Prisotnost poznorimskih prebivalcev označujejo posamezni pomembnejši predmeti iz tistega časa (fibule, pasne spone, jermenski zaključki idr.), predvsem pa številni rimski novci. Vendar nam doslej ni uspelo podrobneje določiti obsega in značaja poznorimske naselbine, ki bi glede na nekatere značilne najdbe moške noše lahko imela vojaški značaj (Osmuk 1997a, 15). Drugo poznorimsko najdišče Molida se nakazuje na zahodnem delu Kobariškega blata, že v bližini zaselka Robič, in je pomembno predvsem zaradi najdbe zlatnika cesarja Valentinijana III. (glej pogl. 1.5). Pri tem arheološkem najdišču je že Matej Župančič domneval povezavo s sistemom Claustra Alpium luliarum (Župančič 1991). Soča Valley. However, this was not the sole Late Roman post in the vicinity of Kobarid. In this area numerous important routes converged. Along these routes and then along the Nadiža river one could enter Italy by bypassing the main entries, so this region became increasingly important. The archaeological research in the area of Gradič above Kobarid has confirmed that during the Roman times a larger settlement with a cult place was created on the old prehistoric site and the slopes below it. The presence of the Late Roman inhabitants is marked by individual important finds from the period (fibulae, belt buckles, strap ends, etc.) and especially by the numerous Roman coins. However, we have so far not managed to define the scope and character of the Late Roman settlement, which could - taking into account some typical male attire - have a military character (Osmuk 1997a, 15). Glede na izredno pomembno strateško mesto najdišča pred vhodom v sotesko, ki ob Nadiži vodi v Furlanijo, se zdi domneva smiselna in sprejemljiva. Ob tem je treba omeniti, da največji del najdenih zlatnikov iz 5. st. kaže na ostanke vojaških plač, saj zakope takšnih novcev velikokrat ugotavljamo na izpostavljenih strateških mestih, kamor so takratni strategi pošiljali manjše vojaške oddelke, sestavljene pretežno iz najetih barbarskih vojakov, federatov (Kos 1986, 223; Werner 1992, 411-417; Božič, Ciglenečki 1995, 262). Strateški pomen območja Molide kaže tudi postavitev zidu prečno čez dolino v času Napoleonovih osvajanj. Zid se na južni strani doline povsem približa postojanki v Molidi. Vendar območje brez raziskav ne dopušča določnejših sklepov. Prav tako slabo je znana rimska naselbina v Starem selu, zato tudi ni razvidna morebitna povezava obeh najdišč, ki ležita v veliki bližini na nasprotnih straneh doline. 5.1.4 POMEN PROMETNEGA VOZLIŠČA PRI KOBARIDU V OBRAMBNI ZASNOVI Tractus Italiae circa Alpes Podroben pregled poti, ki so vodile v Posočje, in poznorimske najdbe na območju Kobarida, ki kažejo na zelo okrepljeno vlogo prometnega vozlišča pred enim izmed najlažjih prehodov iz doline Soče v Furlanijo, dovoljujejo domnevo, da je v poznorimskem času nastal tu sklop naselbin in utrdb, ki so imele nalogo nadzirati premike čez to območje in po potrebi še zadnjič zadržati sovražnika pred vrati Italije. Pri tem je naloga najpomembnejšega branika ceste, ki je vodila čez Predel, pripadla prav utrdbi na Tonovcovem gradu. Ob skicirani sliki poznorimske obrambe se zastavlja več vprašanj kot odgovorov. Predvsem nam še ni uspelo časovno ločiti sistema linearnih zapornih zidov od globinsko zasnovane obrambe, katere močan člen je bilo prav obravnavano območje pri Kobaridu. Vemo, da sta bila vsaj delno sočasna, v poznejšem obdobju, po začetku 5. st., pa kaže, da so linijsko obrambo povsem opustili in okrepili delovanje strateško umeščenih postojank ob pomembnejših prometnicah. To bi nakazovale predvsem posamezne najdbe novcev iz prvih treh desetletij 5. st. v postojankah Tonovcov grad, Most na Soči, Sv. Marija na Jezeru idr. (Ciglenečki, Milavec 2009; glej tudi pogl. 1.5). Nakopičene poznorimske najdbe vojaškega značaja so bile najdene tudi v utrjeni višinski postojanki Castelraimondo ob Tilmentu. Sodijo v fazo 4a, ki so jo datirali v 4. st. in vse do okoli leta 430 (Santoro Bianchi 1992, 185-194). Omeniti je treba tudi zgodovinske vire, ki so prav za ta čas in to območje najzgovornejši (pregledno pri Šašel 1971; Bratož 1994; Marcone 2004, 352-359). Posebej zanimivo je, da se v virih dovolj značilno omenjajo tako claustra, torej umetni zaporni zidovi, kot tudi clausurae, Molida, the second Late Roman site, is indicated in the west part of Kobariško blato, in the vicinity of the hamlet of Robič, and is important because of the find of the gold coin of Emperor Valentinian III (see chapter 1.5). Matej Župančič assumed a connection between this archaeological site and the Claustra Alpium Iuliarum (Župančič 1991). This seems a reasonable assumption if we take into account its important strategic location in front of the entrance into the gorge that runs along the Natisone river and into Friuli. Of course, we have to mention that finds of 5th century gold coins often indicate remains of military salaries as these were often buried on exposed strategic locations to which the strategists of the time sent small military units that were predominantly composed from hired barbarian soldiers, i.e. foederati (Kos 1986, 223; Werner 1992, 411-417; Božič, Ciglenečki 1995, 262). The strategic importance of the Molida area is also shown by the wall that spanned across the valley during the Napoleonic conquests. On the south of the valley the wall almost reached the post in Molida. However, the area would need to be researched further if we wished to draw any definite conclusions. The Roman settlement in Staro selo is still relatively poorly known, thus no connections have been ascertained so far even though the two sites lie just across the valley close to each other. 5.1.4 THE SIGNIFICANCE OF THE CROSSROADS AT KOBARID FOR THE DEFENSIVE SYSTEM Tractus Italiae circa Alpes A detailed overview of the routes that led into the Soča Valley, combined with the Late Roman finds in the Kobarid area - which show a strongly fortified role of the transport crossroads in front of one of the easiest passes from the Soča Valley into Friuli - allow for the assumption that a number of settlements and forts appeared in this area during the Late Roman times. They were built in order to control the movements across this area and to prevent the enemy from setting foot into Italy. The most important defensive post on the road that led across Predel was the fort on Tonovcov grad. The sketched outline of the Late Roman defence poses more questions than answers. As regards the time of origin we have not managed to differ between the system of linear defensive walls and the defence-in-depth system, a strong part of which was represented by the discussed area near Kobarid. We know that at least for a while they were in use at the same time, however, after the beginning of the 5th century it seems that the linear defensive walls were abandoned and that the strategically placed posts along the important routes were strengthened. This was made especially clear by the individual coin finds from the first three decades of the kar večina zgodovinarjev tolmači kot soteske, ožine, prehode, skratka naravno težko prehodna mesta, ki jih je bilo mogoče uspešno zavarovati. Osredotočanje na umetno linearno obrambo je značilno predvsem za celotno 4. st., medtem ko se v 5. st. v prvi vrsti omenjajo clausurae (Christie 1991, 417-418). To se dobro ujema s sliko, ki jo kažejo zgoraj omenjene utrdbe ob pomembnejših komunikacijah. Izpostavili bi predvsem tri: najbolje raziskani Tonovcov grad, Puštal pri Trnju in Ajdovski gradec pri Bohinjski Bistrici. Glede na številne odkrite predmete vojaške noše in strateški položaj bi smeli vsaj za prvi dve domnevati postojanke treh legij, poznanih iz Noticije dignitatum pod imenom Iuliae Alpinae. Treba pa je poudariti, da bo šele podroben študij gradiva iz doslej le nakazanih postojank omogočil uskladitev arheoloških podatkov z viri in pokazal, kje gre za stalno in kje le za občasno uporabljane utrdbe. Kratek pregled cestnih povezav v severozahodni Sloveniji je pokazal, da je prišlo do intenziviranja življenja v utrdbah in ob njih že v 4. st. in da so bile okrepljene posebej v času, ko je bila opuščena glavna magistrala z vzhoda čez Hrušico v Italijo. Po globini razporejene utrdbe so prevzele breme pritiskov, ki jih je prej v veliki meri zadrževal tristopenjski sistem obrambnih zidov in trdnjav med vrhniko in Hrušico. Tonovcov grad je le prva izmed bolje znanih tovrstnih postojank, večino drugih bo v prihodnosti šele treba raziskati, da bi v celoti razumeli zadnje ukrepe rimske države pri hitrem prilagajanju sistema obrambe na najbolj občutljivem braniku Italije. 5th century in the posts of Tonovcov grad, Most na Soči, Sv. Marija na Jezeru, etc. (Ciglenečki, Milavec 2009; see also chapter 1.5). A concentration of Late Roman finds with a military character was also found in the fortified hilltop post of Castelraimondo on Tagliamento. These finds belonged into phase 4a, i.e. the 4th century and as late as 430 AD (Santoro Bianchi 1992, 185-194). We should also mention the historic sources for this period and area (for an overview see Šašel 1971; Bratož 1994; Marcone 2004, 352-359). It is interesting that the sources mention claustra (i.e. linear defence walls) as well as clausurae, which most historians explain as gorges, straits, passes, i.e. naturally hard to pass locations that were easy to defend. The focus on the linear defence was characteristic for the 4th century, while in the 5th century mostly clausurae are mentioned (Christie 1991, 417-418). This fits the image shown by the above mentioned forts along the important communication routes. We would like to emphasise three of them: the best researched Tonovcov grad, Puštal near Trnje and Ajdovski gradec near Bohinjska Bistrica. Taking into account the numerous military attire finds and combining them with the strategic position of the posts, we could (at least for the first two) assume that these were posts for the three legions, known as Iuliae Alpinae in the Notitia dignitatum. We have to emphasise that only a detailed study of the material from the so far discovered posts will lead to the harmonisation of archaeological data with the sources and show which forts were in permanent use and which were only occasionally used. A short overview of the road connections in northwest Slovenia has shown that life within fortifications and in their vicinity became more intensified in the 4th century and that forts were strengthened especially in the period when the main road leading from the east, across Hrušica and into Italy was abandoned. The defence-in-depth system took over the pressure that was previously taken over by the three level defensive system of walls and towers between Vrhnika and Hrušica. Tonovcov grad is merely the first of the better known posts; most of the others still need to be researched if we wish to understand the final measures that were taken by the Roman state in its quick adjustments made to the defensive system in the most sensitive bastion of Italy. 5.2 tonovcov grad v poznoantični poselitveni sliki vzhodnoalpskega in zahodnobalkanskega prostora 5.2 tonovcov grad in the broader late antique settlement pattern of the east alpine and western balkans 5.2.1 VIŠINSKE POZNOANTIČNE POSTOJANKE NA VZHODNOALPSKEM OBMOČJU Za boljše razumevanje kompleksne podobe Tonovcovega gradu v drugi polovici 5. in v celotnem 6. st. je treba najprej predstaviti poznoantične višinske utrdbe kot temeljno značilnost spremenjene poselitvene slike na vzhodnoalpskem območju. Strukturne spremembe v poznoantični družbi, gospodarstvu in vojski, do katerih je prišlo na celotnem območju rimskega imperija, so bile še izrazitejše prav na ozemlju današnje Slovenije, ki je bilo strateško zelo pomembno, saj je mejilo na province Norik, Panonijo in Dalmacijo. Tu je bil najlažji prehod z vzhoda in severa v Italijo in zaradi številnih državljanskih vojn in premikov barbarskih ljudstev proti zahodu tudi najbolj ogrožen (Bratož 2003; Cedilnik 2004). V zadnjih desetletjih opravljene raziskave po-znoantičnih višinskih postojank na tem območju so povsem spremenile prvotno predstavo o enotnem tipu poznoantičnega "refugija", ki je bila dolgo v veljavi (Šašel 1972, 5; 1980, 14; Petru 1972, 357; 1982, 306). Razkrile so širok spekter tipov utrdb, ki ga novejše raziskave sistematično dopolnjujejo (Ciglenečki 1979, 460-466; 1987a, 111-120; 1994a, 240-248; 2008). V drugi polovici 5. in v celotnem 6. st. je bila pestrost oblik višinskih naselbin največja. V veliki meri so bile ponovno izkoriščene prav tiste postojanke, ki so bile poseljene že prej, vendar pa nikjer ni bilo mogoče povsem zanesljivo dokazati kontinuitete med tistimi iz prve polovice 5. st. in naselbinami, ki so bile v celotnem obsegu zgrajene v drugi polovici 5. ali celo v 6. st. (Ciglenečki 1999, 292). Osnovna značilnost postojank zadnjega obdobja je pretežno stalna oziroma dalj časa trajajoča poseljenost. Za bivališča so uporabili na vse načine zaščitene postojanke, od povsem naravno zavarovanih do zelo močno umetno utrjenih. Pri tem so se pokazale razlike v obliki in obsegu poselitve: razlikovati je mogoče naselbine, v katerih so stavbe enakomerno razvrščene po večjem delu notranjosti (Gradec pri Prapretnem, Ajdna, Vipota, Sv. Radegunda, Tonovcov grad, Gradišče nad 5.2.1 LATE ANTIQUE HILLTOP SETTLEMENTS IN THE EAST ALPINE AREA In order to obtain a better understanding of the complex image of Tonovcov grad in the second half of the 5th and throughout the 6th century we have to start with the presentation of the Late Antique hilltop posts as a basic characteristic of the changed settlement pattern in the East Alpine area. The structural changes in the Late Antique society, economy and military forces (which took place across the entire Roman Empire) were even more noticeable in the area of today's Slovenia, which was of significant strategic importance as it was here that Italy bordered on the provinces of Noricum, Pannonia and Dalmatia. This was where it was the easiest to enter Italy from the east and north and the numerous civil wars and movements of barbaric tribes made it also the most endangered area (Bratož 2003; Cedilnik 2004). Over the last few decades the research conducted in the Late Antique hilltop sites in this area has completely changed the original view of the unified type of a Late Antique 'refuge' (Šašel 1972, 5; 1980, 14; Petru 1972, 357; 1982, 306). Research has revealed a broad spectre of fortification types, to which new ones are systematically added with the newer research (Ciglenečki 1979, 460-466; 1987a, 111-120; 1994a, 240-248; 2008). The most diverse shapes of hilltop settlements could be found in the second half of the 5th and throughout the 6th century. To a great extent the previously used settlements were reused, however it is impossible to reliably prove any continuity between the settlements from the first half of the 5th century and those that were constructed in the second half of the 5th or even in the 6th century (Ciglenečki 1999, 292). One of the main characteristics of the posts from the second period was that they were settled for a longer period of time and they were in continuous use. All possible types of protection were used in order to protect the settlements, thus they ranged from entirely naturally protected to extremely artificially fortified. At this certain differences have been noticed in the form Bašljem - Ciglenečki 1987a, 114), in take, pri katerih so osredotočene le na ožje omejeno območje v naselbini in ob njenem obzidju (Rifnik, Vrtovin, Sv. Lovrenc nad Bašljem - Ciglenečki 1987a, 115-116). Poseben tip, ki je bil sprva zelo slabo razviden, so bile postojanke na strateško pomembnih mestih, kjer je bilo mogoče ugotoviti le za obrambo namenjene stavbe in obzidje, v notranjosti pa skromno zgodnjekrščansko cerkev (Ciglenečki 1979, 460). Ta tip utrdbe, ugotovljen najprej na Korinjskem hribu pri Velikem Korinju, se je s poznejšimi raziskavami vedno znova potrjeval kot posebnost v spektru višinskih naselbin (Ciglenečki 1985, 256). Tudi dejstvo, da so med novejšimi raziskavami tega tipa postojank odkrili največ germanskega gradiva, dodatno opozarja na njihov strateški pomen (Božič, Ciglenečki 1995; Klasinc 2001, 67-68; Lux 2006; Ciglenečki 2008, 507-511). Zaokroženo skupino sestavljajo tudi utrjene postojanke z močnim poudarkom na sakralnem delu in pripadajočih bivalnih zgradbah, ki so podrobneje obravnavane na drugem mestu (glej pogl. 4.2). obstoj pribežališč v 6. st. je najslabše znan. Na območju posočja nakazujeta na to, da gre za pribežališče, postojanki Gradec pri Logjeh (sl. 1.28, 1.29) in Gradec pri Drežnici (sl. 1.26,1.27). Njun majhen obseg, odmaknjena lega in doslej ugotovljene skromne najdbe dovoljujejo domnevo, da sta pripadali skupinam ljudi, ki so živeli v tamkajšnji okolici in so se vanju zatekli le v času nevarnosti (glej pogl. 1.5). Tej skupini je mogoče pripisati tudi najdbe na prazgodovinskih gradiščih, na katerih so pri sondiranjih odkrili šibke sledove kratkotrajnega bivanja poznoantičnih prebivalcev (Dular et al. 1991, 69-76; 2000, 127-129; Pavlin, Dular 2007, 66-78). 5.2.2 TEMELJNE ZNAČILNOSTI NASELBINE NA TONOVCOVEM GRADU and scope of the settlement: ranging from settlements in which the buildings are equally spread across the majority of the interior (Gradec near Prapretno, Ajdna, Vipota, Sv. Radegunda, tonovcov grad, Gradišče above Bašelj - Ciglenečki 1987a, 114), to ones at which the buildings are focused only in a small area in the settlement and along its walls (Rifnik, Vrtovin, Sv. Lovrenc above Bašelj - Ciglenečki 1987a, 115-116). A special type that was at first hardly recognisable was represented by posts on strategically important locations where only buildings intended for defensive purposes, the fortification and a modest Early Christian church were found (Ciglenečki 1979, 460). This type of fort was first noticed on Korinjski hrib near Veliki Ko-rinj and was later repeatedly confirmed as a speciality within the hilltop settlements spectre (Ciglenečki 1985, 256). Also the fact that in the new research most of the discovered material of Germanic origin came from these posts, additionally emphasised their strategic importance (Božič, Ciglenečki 1995; Klasinc 2001, 67-68; Lux 2006; Ciglenečki 2008, 507-511). Another group is represented by fortified posts with a strong emphasis on the ecclesiastical part and the accompanying living quarters, described elsewhere in greater detail (see chapter 4.2). The refuges from the 6th century are the least explored. In the area of the Soča river the posts Gradec near Logje (Figs. 1.28, 1.29) and Gradec near Drežnica (Figs. 1.26,1.27) indicate that they were used as refuges. Their small size, remote locations and modest finds lead to the assumption that they belonged to groups of people who lived in the vicinity and used this refuges only in periods of danger (see chapter 1.5). This group also includes finds on prehistoric sites, at which trial trenching revealed modest traces of short-term settling by Late Antique inhabitants (Dular et al. 1991, 69-76; 2000, 127-129; Pavlin, Dular 2007, 66-78). utrjena naselbina na tonovcovem gradu z zidano stavbno arhitekturo, obzidjem s stolpi in zgodnje-krščanskim središčem na vzvišenem mestu je izrazit in pomemben primer postojanke sklepne faze poznoan-tičnih najdišč na vzhodnoalpskem območju in hkrati tudi ena izmed najbolje ohranjenih. treba je izpostaviti predvsem tri vidike, ki določajo njen pomen in vlogo v sočasni poselitveni sliki. ti so: 1. Domišljeno zasnovan tloris naselbine s stavbami, razporejenimi po celotni površini. 2. Geografska umeščenost utrjene naselbine in njena vloga v sistemu tedanjih komunikacij. 3. Kompleksen cerkveni sklop v naselbini. 5.2.2 THE BASIC CHARACTERISTICS Of THE SETTLEMENT AT TONOVCOV GRAD The fortified settlement with stone buildings, defensive walls with towers and an elevated Early Christian centre is an important example of an East Alpine post from the last Late Antique phase; it also happens to be one of the best preserved. Three aspects need to be pointed out, for they define its importance and role in the settlement pattern at the time. These are: 1. A deliberate layout with buildings on the entire area of the settlement. 2. The geographic position of the fortified settlement and its role in the communication system of the time. 3. The complex ecclesiastical centre in the settlement. OBLIKOVANOST NASELBINE IN NJENA TLORISNA ZASNOVA Poznoantična naselbina na Tonovcovem gradu je bila naselbina, zasnovana po enotnem konceptu in zelo domišljeno zgrajena (glej pogl. 1.1.3): skupek cerkva na vzvišenem skalnem platoju in z različno velikimi in različno oblikovanimi stanovanjskimi zgradbami smiselno zapolnjena notranjost. Po tlorisni zasnovi s pet prostori v dveh vrstah izstopa stanovanjski objekt št. 28, ki leži v bližini cerkvenega sklopa in je glede na višino le nekoliko pod njim (sl. 1.7). Njegove vloge brez raziskav ni mogoče zadovoljivo pojasniti. Hipotetično bi ga bilo mogoče povezati z bivališčem duhovščine v naselbini. Primerjava koncepta tlorisne zasnove kot tudi gradnja stavb v naselbini razkrivata močne podobnosti z nekaterimi že raziskanimi najdišči, pa tudi s tistimi, pri katerih je mogoče že zgolj na površini zanesljivo prepoznati obrise stavb. Kot najprepričljivejša analogija se ponuja Gradec pri Prapretnem s sicer majhno, a na gosto pozidano in povsem domišljeno tlorisno zasnovo naselbine (sl. 5.9). To kaže na hkraten nastanek naselbin, kjer so v osrednjem delu razvrščene stavbe posebnega pomena (cerkev, večje stanovanjske zgradbe), na robu pa manjše, od katerih so imele nekatere verjetno tudi vlogo obrambnih stolpov. Drobno gradivo in novci potrjujejo sočasnost z naselbino na Tonovcovem gradu (Ciglenečki 1981; Bausovac 2003). THE SETTLEMENT DESIGN AND LAYOUT The Late Antique settlement at Tonovcov grad shows a carefully planned unified concept (see chapter 1.1.3): it reveals three churches on the elevated rock plateau and an interior filled with various sized and shaped living quarters. The building No. 28, which was used as living quarters, stands out with its five rooms in two rows. The building is located close to the three churches (Fig. 1.7) and is positioned on an only slightly lower elevation. Further research would be necessary if we wished to explain its role; however, it could hypothetically be linked to the living quarters of the clergy. The comparison of the ground plan and building techniques used in the settlement reveal great similarities with some previously researched sites as well as with those at which it was possible to reliably recognise the outlines of the buildings already from the surface. The most convincing analogy is that with Gradec near Prapretno which has a small, but densely constructed and excellently thought out settlement ground plan (Fig. 5.9). This indicates the settlement was built in a single event, in which the buildings with special importance (church, larger buildings with living quarters) were placed in the centre while smaller buildings - some of which are believed to have had also the role of defensive towers - were placed on the edge. The small finds and coins confirm the simultaneous- ® 20 m Sl. 5. 9: Gradec pri Prapretnem. Utrjena poznoantična naselbina (Ciglenečki 1987a, 54). Fig. 5.9: Gradec near Prapretno. Late Antique fortified settlement (Ciglenečki 1987a, 54). Sl. 5.10: Ajdna nad Potoki. Poznoantična naselbina (Sagadin 1997, na ovitku). Fig. 5.10: Ajdna above Potoki. Late Antique settlement (Sagadin 1997, on cover). Nedvomno enak značaj kaže poznoantična naselbina Ajdna nad Potoki (sl. 5.10), kjer je bil ves razpoložljivi prostor v celoti zapolnjen s cerkvijo in bivalnimi zgradbami, nepozidan je ostal le vrh s koto 1.048 m, saj je močno izpostavljen vetrovom (Leben, Valič 1978; Sagadin 1997). Sorodno naselbino bi smeli videti v postojanki Kapelle pri Jadersdorfu (sl. 5.11). tu so dosedanje raziskave odkrile le nekaj tlorisov stavb, vendar razgibana površina na odlično zavarovanem skalnem čoku nakazuje po celi površini razporejene zgradbe, pri čemer gre morda tudi za kombinacijo zidane in lesene arhitekture (Felgenhauer - Schmiedt 1993). Ni povsem jasno, ali smemo primerljiv tip naselbine domnevati na Šenturški gori (Ulrichsberg), kjer so na najvišjem delu naselbine raziskali strnjeno razporejene bivalne zgradbe in cerkev (sl. 5.12). Spričo ne povsem zanesljive datacije nekaterih zgradb nižje v naselbini in ne v celoti rešenega vprašanja lokacije obrambnega zidu med obema najdiščema ni mogoče potegniti zanesljivih vzporednic (Egger 1950; Glaser 1996b, 59-60, sl. 24). Pri Raveu v Karniji, na še povsem neraziskanem griču Cuel Budin, je mogoče v manjši naselbini v reliefu zaznati nekaj poznoantičnih zgradb, ki so razporejene ness with the settlement at Tonovcov grad (Ciglenečki 1981; Bausovac 2003). The Late Antique settlement Ajdna above Potoki shows a similar character (Fig. 5.10). Here, all of the available area was filled with the church and living quarters, while only the peak at 1048 m.a.s.l. remained free of any constructions as it was exposed to strong winds (Leben, Valič 1978; Sagadin 1997). A similar settlement was discovered at Kapelle near Jadersdorf (Fig. 5.11). The research conducted so far revealed merely a few ground plans; however the shape of the surface on the excellently protected rock pinnacle indicates buildings that were spread across the entire area, at which it is possible that we are dealing with a combination of stone and wooden architecture (Felgenhauer-Schmiedt 1993). It is unclear whether we can assume a similar settlement type on Ulrichsberg, where closely lined up living quarters and a church have been researched at the highest point of the settlement (Fig. 5.12). Due to the unreliable dates of some buildings lower down in the settlement and the unsolved issue as to the location of the defensive wall it is impossible to draw reliable parallels (Egger 1950; Glaser 1996b, 59-60, Fig. 24). Sl. 5.11: Kappele pri Jadersdorfu. Utrjena poznoantična naselbina (po Felgenhauer - Schmiedt 1993, t. 2). Fig. 5.11: Kapelle near Jadersdorf. Late Antique fortified settlement (after Felgenhauer-Schmiedt 1993, Pl. 2). One can recognise a few Late Antique buildings in the relief within a smaller settlement on a hill called Cuel Budin (that has not been researched yet) near Raveo in Carnia where the buildings are scattered across most of the interior. Through the similar longitudinal buildings with multiple rooms and the finds discovered so far, this site confirms contemporaneity and partial similarity with tonovcov grad (Rupel 1997; Villa 2001, 859). Some excellent analogies have been ascertained while comparing tonovcev grad with sites in the more distant Balkan area. One of the best analogies is represented by Golemanovo Kale in Bulgaria (built in the time of justinian I; Fig. 5.13), which was built according to a unified plan with densely positioned houses (Werner 1992, 396-400). Newer research has also shown great similarities with the large fortified settlement on jelica in serbia (Fig. 5.14), discovered and researched during the last decades (Milinkovic 2010). There are numerous comparisons in the well-researched Macedonia, e.g. Golemo Gradiste near Konjuh, Davina Kula near Cucer (Fig. Sl. 5.12: Šenturška gora (Ulrichsberg). Poznoantična naselbina (po Glaser 1996b, sl. 24). Fig. 5.12: Ulrichsberg. Late Antique settlement (after Glaser 1996b, Fig. 24). po večjem delu notranjosti. S podobno oblikovanimi podolgovatimi zgradbami z več prostori in dosedanjimi najdbami potrjuje sočasnost in delno sorodnost s to-novcovim gradom (Rupel 1997; Villa 2001, 859). V primerjavah z bolj oddaljenim balkanskim območjem je mogoče najti nekaj odličnih analogij za tovrstno obliko naselbine. takšna je predvsem v celoti raziskana Golemanovo Kale v Bolgariji (sl. 5.13), zgrajena 5.15), Gradiste near Dramce, Manastir near Sokolarci (Mikulcic 2002, 128-153, 153-158, 206-207, 232). taking into account the layout of the buildings we can also compare the ground plans of settlements in which only wooden buildings or wooden buildings on dry walls have been discovered so far. Such is for instance the excellently researched Colle Santino (Fig. 5.16) in Invillino (Bierbrauer 1987), while resume excavations Sl. 5.13: Sadovec. Utrjena poznoantična naselbina Golemanovo Kale (po Uenze 1992, 557). Fig. 5.13: Sadovec. Late Antique fortified settlement Golemanovo Kale (after Uenze 1992, 557). po enotni zasnovi na gosto razporejenih hiš, ki je nastala v justinijanskem obdobju (Werner 1992, 396-400). Z novejšimi raziskavami kaže velike podobnosti tudi v zadnjih desetletjih odkrita velika utrjena naselbina na Jelici (sl. 5.14) v Srbiji (Milinkovič 2010). Številne so primerjave iz dobro raziskane Makedonije, npr. Golemo Gradište pri Konjuhu, Davina Kula pri Čučerju (sl. 5.15), Gradište pri Dramču, Manastir pri Sokolarcih (Mikulčič 2002, 128-153, 153-158, 206-207, 232). Glede razporejenosti stavb lahko primerjamo tudi tlorisne zasnove naselbin, kjer so doslej ugotovili le povsem lesene zgradbe ali lesene zgradbe na temeljih iz suhega zidu. Takšen je v celoti raziskan Colle Santino (sl. 5.16) v Invillinu (Bierbrauer 1987), zaščitne raziskave in Sl. 5.14: Jelica. Utrjena poznoantična naselbina (Milinkovič 2010, sl. 8) Fig. 5.14: Jelica. Late Antique fortified settlement (Milinkovič 2010, Fig. površinski pregledi pa tovrstne oblike naselbin nakazujejo tudi na Svetih gorah na Bizeljskem (Ciglenečki 1992b, 67-70), Gori pri Polhovem Gradcu (Slabe 1982-1983, 68; sl. 5.17) in Tinju nad Loko pri Žusmu (Ciglenečki 2000; sl. 5.18). pri omenjenih primerjavah ni mogoče vedno zanesljivo sklepati na funkcijo tako koncipiranih naselbin. Ugotavljamo, da ta tip naselbine najbrž ni nastal izključno iz strateških razlogov, in domnevamo, da je šlo predvsem za osrednja naselbinska jedra nekaterih območij. primerjave bi tako kazale, da je šlo pri Tonov- Sl. 5.15: Čučer. Utrjena poznoantična naselbina Davina Kula (Mikulčic 2002, sl. 45). Fig. 5.15: Čučer. Late Antique fortified settlement Davina Kula (Mikulčic 2002, Fig. 45). and surface surveys indicate similar settlement forms on Svete gore in Bizeljsko (Ciglenečki 1992b, 67-70), Gora near Polhov Gradec (Slabe 1982-1983, 68; Fig. 5.17) and Tinje above Loka pri Žusmu (Ciglenečki 2000; Fig. 5.18). In the aforementioned comparisons it is not always possible to reliably conclude the function of these settlements. We are of the opinion that this type of settlement was not developed exclusively for strategic reasons, and we assume that they represented a central settlement in certain areas. Comparisons show that Tonovcov grad had an emphasised role of a large, solidly built settlement, which could have represented the centre of a smaller region. It seems that the larger settlements of this type were some sort of miniature towns from the last Antique period, and in some cases also centres of the last organised territorial units. THE IMPORTANCE OF THE STRATEGIC POSITION OF THE SETTLEMENT AT TONOVCOV GRAD The second important aspect of the settlement at Tonovcov grad is represented by the geographic location of the post, especially its position alongside the important route leading from Carinthia, across the Predel/Pre-dil pass and along the Soča river valley into Friuli (Fig. 5.1). As indicated in the overview of the Late Roman problematic (see chapter 5.1), the shortest route between Friuli and Carinthia gained on importance in the 4th and during the first half of the 5th century, when beside the itinerary road across Hrušica a number of other routes came into intensive use. These routes made the access to Italy easier and enabled manoeuvres of military divisions through the broader area. As revealed by the precisely dated objects and coins from Tonovcov grad and other posts, the route Carinthia-Predel-Cividale del Friuli preserved its importance into the 5th and 6th centuries. Some of this importance was preserved long into the Middle Ages and partially also into the Modern times (chapter 1.6; Bosio 1957, 36-38; 1991, 193-194; Gestrin 1987; Šumrada 1987; Holz 1994, 25). It seems that the abandonment or at least poorer maintenance of the main Roman roads and a different military doctrine deriving from the Diocletian's reforms were responsible for this. Theoderic's proclamation addressed at the Lucristans at Soča in a way deals with the issues of the roads in this area. This proclamation addresses the negligence in road maintenance and exposes the need to renew this system (Cassiodor, Var., I. 29). There are two possible explanations behind the name supra Sontium. The first Sl. 5.16: Colle Santino, Invillino. Utrjena poznoantična naselbina (po Bierbrauer 1987, pril. 2). Fig. 5.16: Colle Santino, Invillino. Late Antique fortified settlement (po Bierbrauer 1987, Insert 2). Sl. 5.17: Gora nad Polhovim Gradcem. Utrjena poznoantična naselbina (Slabe 2009, 268-269). Fig. 5.17: Gora above Polhov Gradec. Late Antique fortified settlement (Slabe 2009, 268-269). /////// //i /// //, /// // ////A // / / /// /C /// /// -\UUiy! ————^ / // / / //, lil lil w (d J^ka pri _ 25 m Si. 5.18: Tinje nad Loko pri Žusmu. Poznoantična naselbina (po Ciglenečki 2000, sl. 9). Fig. 5.18: Tinje above Loka pri Žusmu. Late Antique settlement (after Ciglenečki 2000, Fig. 9). covem gradu za poudarjeno vlogo večje, solidno grajene naselbine, ki je morda predstavljala tudi središče manjše regije. Pri večjih naselbinah tega tipa kaže, da gre za nekakšna miniaturna mesta iz zadnjega obdobja antike, v katerih je mogoče videti ponekod tudi središča zadnjih organiziranih teritorialnih enot. is the explanation given by Fran Kos, who assumed the Goths were living in the Soča river valley, while the other was given by Jaro Šašel (1975, 80) and backed by Rajko Bratož (2000, 45-46), both of whom understood that this syntax was used to denote the Vipava Valley. It was the research at Tonovcov grad with the rich POMEN StRAtESKE LEGE NASELBINE NA tONOVCOVEM GRADU Drugi pomemben vidik naselbine na tonovcovem gradu predstavlja geografska umeščenost postojanke, predvsem njena lega ob pomembni vpadnici iz Koroške čez prelaz Predel (sl. 5.1) in po dolini Soče v Furlanijo. Kot smo nakazali že pri pregledni predstavitvi poznorimske problematike (glej pogl. 5.1), je najkrajša povezava med Furlanijo in Koroško pridobila pomen že v 4. st. in v prvi polovici 5. st., ko so poleg itinerarske ceste čez Hrušico intenzivneje zaživele tudi druge poti, ki so olajšale dostop v Italijo in omogočale manevriranje vojaških oddelkov na širšem območju zapornega sistema. Pot Koroška-Predel-Čedad je pomen ohranila tudi v 5. in 6. st., kot pričajo dobro datirani predmeti in novci s tonovcovega gradu ter iz nekaterih drugih postojank vzdolž nje, in ga zadržala globoko v srednji in delno celo v novi vek (Bosio 1957, 36-38; 1991, 193-199; Gestrin 1987; Sumrada 1987; Holz 1994, 25). Zdi se, da sta temu botrovali predvsem opustitev ali slabše vzdrževanje glavnih rimskih cest in drugačna vojaška doktrina po dioklecijanskih reformah. Posredno se na problematiko cest na tem območju nanaša prav teoderikov razglas Lukristanom ob Soči, ki govori o zanemarjanju vzdrževanja cest in potrebi po obnovi tega sistema (Cassiodor, Var., I. 29). O oznaki supra Sontium obstajata dve možni razlagi: prvo je zapisal že Fran Kos, ki je domneval gotske prebivalce v Posočju, drugo pa Jaro Sašel (1975, 80), ki je pod sintagmo supra Sontium domneval območje Vipavske doline. V tej razlagi se mu je pridružil tudi Rajko Bratož (2000, 45-46). Prav raziskave na tonovcovem gradu pa so z bogatimi sledovi iz vzhodnogotskega obdobja (tudi pet vzhodnogotskih srebrnikov; glej tonovcov grad. Najdbe, pogl. 5.1, kat. št. 158-162) obrnile pozornost spet k starejši Kosovi tezi, ki tudi neposredneje ustreza zgoraj omenjeni označbi. Pri tem je treba opozoriti, da so bili vzhodnogotski novci doslej najdeni predvsem v severnem delu zahodne Slovenije (Kranj, Puštal, Ajdna), niso pa poznani iz Vipavske doline. Seveda s tem nikakor ne zanikamo povezave po Vipavski dolini, ki je naravno ena najugodnejših, opozarjamo pa na možnost, da so Lukristani prebivali tudi ob Soči in vzdrževali cesto, ki je po dolini Soče vodila na Koroško (Ciglenečki 1994b, 8; Ciglenečki 2008, 522). O mreži cest, ki smo jo poizkusili rekonstruirati za poznorimsko obdobje na območju alpskih zapornih zidov (glej pogl. 5.1), smemo z veliko verjetnostjo - čeprav morda v nekoliko manjšem obsegu - domnevati, da so jo uporabljali tudi v drugi polovici 5. st. in v 6. st. Poleg že opisanih cest je treba omeniti tudi cesto, ki je iz Akvileje vodila proti severu: to je dobro znana itine-rarska cesta Akvileja-Virunum, ki jo označujejo miljniki (Bosio 1970, 147-177; 1991, 157-171). Pomembnost ceste, ki je iz Čedada vodila čez Predel na Koroško in Eastern Gothic period traces (including five Eastern Gothic silver coins; see Tonovcov grad. finds, chapter 5.1, Cat. No. 158-162) that turned the attention once again to the older thesis by Kos. Until this discovery Eastern Gothic coins were mostly found in northwest Slovenia (Kranj, Puštal, Ajdna), and not a single find came from the Vipava Valley. Of course this by no means denies the connections across the Vipava Valley, which represents one of the most favourable routes, however we would like to draw attention to the possibility that the Lucristans also lived along the Soča river and maintained the road which lead along the Soča river valley and towards Carinthia (Ciglenečki 1994b, 8; Ciglenečki 2008, 522). We tried to reconstruct the Late Roman road network in the Alpine area (see chapter 5.1), and we can, with great certainty assume that it was used also in the second half of the 5th as well as the 6th century, even though possibly in a slightly reduced scope. Apart from the previously described roads we also have to mention the road that lead north from Aquileia: this is the well-known itinerary road Aquileia-Virunum that is marked by milestones (Bosio 1970, 147-177; 1991, 157-171). The importance of the road that lead from Cividale del Friuli, past Tonovcov grad and across Predel towards Carinthia can be understood through the more important role of Cividale del Friuli in Late Antiquity: the network of routes that lead towards Italy from the east and north, ended up at the Soča river, where the transport was directed along the valley of Nadiža/Natisone through Cividale del Friuli and towards Aquileia and Ravenna. A connecting road that joined the itinerary road towards Virunum at Gemona ran along the foothills of Colli from Cividale del Friuli. Recently the research carried out alongside this route has shed some light on numerous Late Antique forts, some of which were located directly at the roadside, while others were moved slightly away. The first group includes Artegna and Gemona, while the second group includes the posts San Giorgio di Attimis and Monte Zuccon near Nimis (Villa 2001, 847-848, 852-855; 2006; Cipollone 2006). Recently traces of a strong fortification wall with towers, a cistern and a rich settlement layer that documents the events in the Late Antiquity were discovered at Artegna (excavations carried out by Luca Villa in 2007, unpublished). At the entrance to the Val Canale in the Gailtal Valley, where the itinerary and the Soča river valley road met on what is today the Austrian side, an important role was played by the post on Hoischhugel (Figs. 5.1, 5.19), located on a naturally protected position with great strategic importance (Egger 1916, 93-104; Ciglenečki 1987a, 47-48; Glaser 1996b, 69-72). The special importance of certain Late Antique posts in the East Alpine area was indicated already with the discovery of Korinjski hrib (Fig. 5.20): the special type of fortification with an emphasised strategic im- Sl. 5.19: Hoischügel (Hanjžev hrib). Poznoantična utrdba (Glaser 1996b, sl. 31). Fig. 5.19: Hoischhügel. Late Antique fort (Glaser 1996b, Fig. 31). tekla mimo Tonovcovega gradu, je mogoče razumeti posredno z močno okrepljeno vlogo Čedada v pozni antiki: mreža poti, ki je z vzhoda in severa vodila proti Italiji, se je namreč iztekla ob Soči, kjer se je promet po dolini Nadiže skozi Čedad usmeril proti Akvileji in Raveni. Iz Čedada je ob vznožju gričevnatega sveta (Colli) vodila povezovalna cesta, ki se je pri Gemoni priključila na itinerarsko cesto proti Virunu. V zadnjem času so raziskovanja ob njej bolje pojasnila kar nekaj poznoantičnih utrdb, ki so deloma stale povsem ob cesti, deloma pa so bile umaknjene nekoliko stran od nje. K prvim je mogoče prišteti Artegno in Gemono, k drugim pa postojanki San Giorgio di Attimis in Monte Zuccon pri Nimisu (Villa 2001, 847-848, 852-855; 2006; Cipollone 2006). Pred nedavnim so pri Artegni odkrili sledove močnega obzidja, utrjenega s stolpi, cisterno in bogato naselbinsko plast, ki dokumentira dogajanje v poznoantičnem obdobju (izkopavanja Luca Ville v letu 2007, neobjavljeno). Na območju Ziljske doline, kjer sta se itinerarska in obsoška cesta na danes avstrijski strani združili, je imela pomembno vlogo postojanka na Hanjževem hribu (Ho-ischhugel), na odlično naravno zavarovani legi in hkrati na izjemnem strateškem kraju (sl. 5.1, 5.19), ob vhodu v Kanalsko dolino (Egger 1916, 93-104; Ciglenečki 1987a, 47-48; Glaser 1996b, 69-72). portance became clear, and it was only confirmed by the newer research of the forts with similar characteristics (for an overview see Ciglenečki 2008, 500-502). The research at Tonovcov grad has shown quite a few elements that indicate the presence of small foreign ethnic groups, thus we wondered, whether special characteristics that would indicate an emphasised military and strategic character of the settlement also exist in the defensive architecture. So far no research has been conducted on the fortification walls and buildings alongside them, thus it is only possible to review the defensive elements on the basis of the stone structures visible on the surface (Fig. 1.7). The defensive wall that in the more visible parts measures 80 cm in width, was fortified by a number of buildings (especially on the easier accessible side) that leaned upon the wall and thus it is possible that some of them served as defensive towers. In the interior the entrance was additionally fortified by a building with three rooms (building No. 22, Fig. 1.7), of which - judging by the height of the ruins - the central one was the highest. The remains of a tower are visible on the east (building No. 31) and most likely also on the northwest, where a large chunk of a wall (building No. 18) can be seen in the corner. Considering the exposed position of this spot a tower was more than likely. (d 20 m Sl. 5.20: Korinjski hrib nad Velikim Korinjem. Poznoantična utrdba (po Ciglenečki 1985, sl. 11). Fig. 5.20: Korinjski hrib above Veliki Korinj. Late Antique fort (after Ciglenečki 1985, Fig. 11). Poseben pomen nekaterih poznoantičnih postojank na vzhodnoalpskem območju smo nakazali že ob odkritju Korinjskega hriba (sl. 5.20): izkristaliziral se je poseben tip utrdbe s poudarjenim strateškim pomenom, ki se potrjuje z novejšimi raziskovanji utrdb s podobnimi značilnostmi (pregledno pri Ciglenečki 2008, 500-502). Raziskave na Tonovcovem gradu so pokazale kar nekaj elementov, ki nakazujejo prisotnost manjših tujih etničnih skupin, zato smo se vprašali, ali obstajajo tudi posebnosti v obrambni arhitekturi, ki bi kazale na poudarjen vojaškostrateški značaj naselbine. Raziskave obzidja in stavb ob njem še niso bile opravljene, zato je mogoče presojati obrambne elemente le na podlagi površinskih opažanj v reliefu dobro vidnih kamnitih struktur (sl. 1.7). Obrambni zid, ki kaže na bolje razvidnih delih širino 80 cm, je bil posebej na lažje dostopni strani okrepljen z mnogimi stavbami, ki so bile prislonjene na zid. Nekatere med njimi so zato lahko služile kot obrambni stolp. Vhod je bil na notranji strani posebej okrepljen s stavbo s tremi prostori (objekt št. 22, sl. 1.7), od katerih je bil - sodeč po ruševini - srednji najvišji. Ostanki ruševin stolpa so vidni na vzhodni strani (objekt št. 31) in zelo verjetno v severozahodnem vogalu, kjer je vidna večja gmota zidu (objekt št. 18). Glede na izpostavljenost tega mesta je obstoj stolpa več kot verjeten. An excellent addition to the defensive system was represented by the wall that was attached to the settlement's defensive wall and which protected the lower lying space on the northwest (Fig. 1.7). It could have served as an outer wall, for it made the approach into the easiest accessible part below the settlement harder. The buildings alongside the wall have a good parallel in the fort on Duel (Fig. 5.21), where the walls were in the second phase strengthened by towers and numerous buildings that were added alongside the wall (Petrikovits 1985, 231). Similar proteichismata were discovered in Bosnia, in the forts of Kostur near Dabrac (Fig. 5.22) and Podgradina Kamenska, both of which were situated alongside important communication routes (Basler 1972, 50-51, 56-57; 1993, 32, 36). The outer wall at the settlement Golemanovo Kale shows a similar distance from the main walls, and in this case it has also been assumed that it had the function of a proteichisma (Werner 1992, 395). A similar proteichisma was discovered in the Byzantine fort of Sant' Antonino in Liguria (Murialdo 1992, 288). Amongst the Slovenian sites we have to mention the certain similarity with the fortified settlement of Rifnik, where defensive towers were added to the walls and a small cluster of solidly constructed stone buildings was Sl. 5.21: Duel. Poznoantična utrdba (Glaser 1996b, sl. 35). Sl. 5.21: Duel. Late Antique fort (Glaser 1996b, Fig. 35). Posebno kvalitetno dopolnilo obrambe predstavlja prečni zid, ki se je navezoval na obrambni zid naselbine in na severozahodni strani varoval nižje ležeči predpro-stor (sl. 1.7). Glede na majhno oddaljenost od glavnega obrambnega zidu je lahko služil tudi kot proteihizma, saj je v veliki meri oteževal dostop v najlažje dostopni del pod naselbino. Osredotočenost zgradb ob obzidju ima dobro vzporednico v utrdbi na Duelu (sl. 5.21), kjer so v drugi fazi okrepili obzidje s stolpi in prizidali ob zid številne stavbe (Petrikovits 1985, 231). Podoben predprostor so zasledili v Bosni pri utrdbah Koštur pri Dabracu (sl. 5.22) in Podgradina Kamenska, ki obe ležita ob pomembnih komunikacijah (Basler 1972, 50-51, 56-57; 1993, 32, 36). Podobno razdaljo od glavnega obzidja ima obzidje predprostora pri naselbini Golemanovo Kale, kjer v njem prav tako domnevajo funkcijo proteihizme (Werner 1992, 395). V Liguriji so podoben predprostor ugotovili na bizantinski utrdbi Sant' Antonino (Murialdo 1992, 288). Med slovenskimi najdišči je treba omeniti tudi določeno podobnost z utrjenim naseljem na Rifni-ku, kjer so v zadnji fazi v obzidje vgradili obrambne stolpe, hkrati pa so v zavetju obzidja postavili manjši skupek solidno grajenih zidanih stavb (Pirkmajer 1994, 46-47). Glede na lego ob pomembni cesti, vojaške najdbe in elemente obrambne arhitekture, ki imajo ustrezne primerjave drugje, se Tonovcov grad kaže kot ena izmed pomembnih utrdb, ki je kontrolirala dostop s severa proti Čedadu in dalje proti Raveni. built just behind the walls, both in the last phase of the settlement (Pirkmajer 1994, 46-47). taking into account its position alongside an important road, the military finds and the elements of defensive architecture that have similar comparisons elsewhere, tonovcov grad appears as an important fort that controlled the access from the north towards Civi-dale del Friuli and further towards Ravenna. THE ECCLESIASTICAL COMPLEX WITHIN THE SETTLEMENT The existence of three Early Christian churches, a supposed baptistery, 'memoria' and an additional unearthed church (building No. 8, Fig. 1.7) in the vicinity emphasises the special importance of the Late Antique settlement of Tonovcov grad (see chapter 4.2). The analysis of East Alpine fortified hilltop settlements with two churches, joined by the specific living quarters and representational buildings, has drawn attention to the importance and uniqueness of such sites within the Late Antique settlement pattern (Ciglenečki 1995, 179-190). The researched double churches in Late Antique fortified hilltop posts in the Eastern Alps have shown a few possible solutions as to their use, however they are most easily explained as regional ecclesiastical centres, for they are exceptionally rare inside Antique town territories (see chapter 4.2). Some sites could be used as pilgrimage stops, however the analysis of the accompa- Sl. 5.22: Kostur pri Dabrici. Poznoanticna utrdba (po Easier 1972, si. 20). Fig. 5.22: Kostur near Dabrica. Late Antique fort (after Easier 1972, fig. 20). CERKVENI SKLOP V NASELBINI obstoj treh zgodnjekrščanskih cerkva z domnevnim baptisterijem, "memorijo" in še eno neizkopano cerkvijo v njihovi neposredni bližini (stavba št. 8, sl. 1.7) poudarja poseben pomen sakralnih objektov v poznoantični naselbini tonovcov grad (glej pogl. 4.2). Analiza najdišč z dvojnimi cerkvami v višinskih utrjenih naselbinah vzhodnoalpskega prostora skupaj s specifičnimi pripadajočimi bivalnimi in reprezentančnimi stavbami je opozorila na izjemnost tovrstnih najdišč v poznoantični naselbinski sliki (Ciglenečki 1995, 179-190). Raziskane dvojne cerkve v višinskih utrjenih poznoantičnih postojankah v vzhodnih Alpah so nakazale več možnih rešitev označitve njihove funkcije, vendar jih je mogoče najustrezneje tolmačiti kot regionalna cerkvena središča, saj so izjemno redke znotraj antičnih mestnih ozemelj (glej pogl. 4.2). Pri nekaterih najdiščih tudi ni mogoče izključiti romarske namembnosti, čeprav ta pri večini glede na rezultate analize pripadajočih bivalnih zgradb ni verjetna (Glaser 1991, 80-85; 2003, 872-879; Ciglenečki 1995, 188-189). Celovit sklop cerkva na tonovcovem gradu je mogoče primerjalno uvrstiti med večja sakralna središča ugašajoče antike na vzhodnoalpskem območju. Predstavlja osrednji verski objekt z več liturgičnimi funkcijami, ki ga je lahko uporabljalo občasno tudi prebivalstvo sosednjih manjših naselbin. Glede na bližino pomembne ceste in ugotovljene relikvije v cerkvi smemo morda domnevati tudi njihovo delno romarsko nying living quarters seem to indicate that this was not the case with most of them (Glaser 1991, 80-85; 2003, 872-879; Ciglenečki 1995, 188-189). The church complex at tonovcov grad can be placed amongst the larger Eastern Alpine ecclesiastical centres from towards the end of Antiquity. It represents a central religious building complex with multiple liturgy functions that could occasionally be used also by the inhabitants of the smaller surrounding settlements. taking into account the vicinity of the important transport route and the relics that have been ascertained in the church we could assume that it was to some extent also used as a pilgrimage stop. It cannot be ascertained whether this was where the bishop from the nearby town of forum lulii occasionally stayed, however this is certainly a possibility. 5.2.3 tHE FINAL CONCLUSIONS as REGARDS tHE LAtE ANtlQUE SEttLEMENt the described aspects of the settlement on tonovcov grad indicate that this was one of the rare posts that combined three important characteristics most commonly only individually present in the typological division of the Late Antique fortified hilltop posts. The comparisons with similar contemporary sites in the East Alpine and West Balkans area show that tonovcov grad was one of the most complex and important Late Antique centres in this area. namembnost. Ali je tu občasno prebival škof iz bližnjega mesta Forum Iulii ali pribežal od kod drugod, ni mogoče zanesljivo ugotoviti, je pa vsekakor ena izmed možnosti. 5.2.3 SKLEPNE UGOTOVITVE O POZNOANTIČNI NASELBINI Opisani trije vidiki naselbine na Tonovcovem gradu kažejo, da gre za eno izmed redkih postojank, kjer so združene kar tri pomembne značilnosti, ki so sicer največkrat prisotne le posamezno v tipološki razčlenitvi utrjenih višinskih poznoantičnih postojank. Predstavljene primerjave s sočasnimi sorodnimi najdišči na vzhodnoalpskem in zahodnobalkanskem območju kažejo, da gre na Tonovcovem gradu za eno izmed najkompleksnejših in najpomembnejših središč zadnjega obdobja antike na tem območju. Zaradi svoje odlične naravne obrambne lege in ne velike relativne višine je bila postojanka primerna za osrednjo naselbino Gornjega Posočja, hkrati pa je postopno dograjevala tudi celovit sklop cerkva, ki je predstavljal sakralno središče regije in imel delno tudi romarski značaj. Strateškoobrambna lega v soteski ob pomembni prometnici je omogočala posadki kontrolo prometa nad vpadnico proti Italiji. Posamezni najdeni barbarski predmeti nam v primerjavi z analognimi utrdbami nakazujejo občasno prisotnost manjših germanskih posadk v naselbini, predvsem vzhodnogotske in langobardske, kar dodatno podčrtuje njen pomen. Naštetim elementom je mogoče pridružiti tudi funkcijo pribežališča, ki jo arheološko na naselbini sicer težko potrdimo, posredno pa jo nakazuje nepozidan utrjen prostor pred naselbino. Utrjena naselbina na Tonovcovem gradu pri Kobaridu je združevala več funkcij in predstavlja eno izmed naselbinskih žarišč, ki mu je v veliki meri uspelo zadržati prvine antične civilizacije z znaki občasne prisotnosti germanskih elementov vse do začetka 7. st. V spremenjeni naselbinski strukturi 6. st. ji je tako mogoče pripisati vlogo, kakršno so imela novonastala transformirana antična mesta. Na širšem območju nekdanje prefekture Ilirik so ugotovili kar nekaj takšnih novonastalih mest, za katera so značilne podobne lastnosti, predvsem pretehtana in močna obramba, v notranjosti pa ob skromnejših bivalnih zgradbah predvsem številne cerkve. Takšni so dobro raziskana Jelica v Srbiji (Milinkovič 2001; 2010), Davina Kula pri Čučerju in Markovi Kuli pri Vodnem v Makedoniji (Mikulčič 2002, 153-158; 190-195) ter na dalmatinskem območju nekaj novonastalih središč v obalnem pasu (Katič 2003). Anonimni Ravenat je naštel nekaj tovrstnih civitates tudi na jugovzhodnoalpskem območju, posebej ob istrski obali in v deželi Carneoli, kjer pa z izjemo Kranja (Carnium) drugih civitates še ni bilo mogoče identificirati (Wolff 2000). Due to its excellent natural defensive position and the not too high relative height, the post could be used as the central settlement in the Upper Soča Valley area. At the same time the church complex was gradually being built and this turned the settlement into a religious centre and to a certain extent also into a pilgrimage stop. The strategic defensive position in the gorge along an important route enabled the military garrison to control the traffic towards Italy. When compared to similar forts the individual Barbarian objects show a contemporary presence of small Germanic garrisons, especially Eastern Goths and Lombards, and this additionally underlines its importance. Even though it is difficult to archeologically confirm this, the settlement could have also functioned as a refuge, as indirectly indicated by the empty fortified area in front of the settlement. The fortified settlement on Tonovcov grad near Kobarid merged multiple functions and represented one of the settlement centres that had managed to preserve elements of Antiquity with traces of occasional Germanic presence until the beginning of the 7th century. In the changed settlement pattern characteristic of the 6th century it can be ascribed the same role as the newly emerging and transformed Antique towns. Quite a few newly formed towns with similar characteristics (a well thought out and strong defensive system, modest living quarters in the interior, numerous churches) have been ascertained in the broader area of the former prefecture of Illyricum: such are for instance the well-researched Jelica in Serbia (Milinkovič 2001; 2010), Davina Kula near Čučer and Markova Kula near Vodno in Macedonia (Mikulčič 2002, 153-158; 190-195) as well as a few newly formed centres on the Dalmatian coast (Katič 2003). The Anonymous from Ravenna named a few similar civitates also in the southeast Alpine area, especially along the Istrian coast and in the land of Carniola - in which Kranj (Carnium) was the only civitas identified so far (Wolff 2000). 6. sklep 6. conclusion V zgornjem delu doline Soče odkrito arheološko najdišče Tonovcov grad je bilo obljudeno periodično vse od prazgodovine do srednjega veka (glej pogl. 2.2). Posebno pomembno je bilo v poznoantičnem času, ko se je tu razvila ena izmed največjih in zelo kompleksnih utrjenih naselbin nadregionalnega pomena. Raziskave hriba, ki s prekinitvami potekajo od leta 1993 dalje (glej pogl. 1.4), so pokazale na obstoj poznoantične naselbine, ki je imela približno 30 zgradb in bila na lažje dostopnih mestih obdana z obrambnim zidom in stolpi (glej pogl. 1.3). Na severozahodni strani je bil tudi večji, s strminami in obzidjem utrjen prostor pod naselbino, ki je lahko služil kot dodatno zavarovanje naselbine ali kot pribežališče v primeru potrebe. Utrjena naselbina obsega vrh izolirane skalne kope v najožjem delu soteske Soče in meri 150 x 90 m. obrambni zid debeline 80 cm je posebej dobro viden na severovzhodni strani, kjer ločuje naselbino od nekoliko nižjega vzhodnega dela hriba. V notranjosti so opazne ruševine stavb, katerih tlorisi so na površini tako dobro vidni, da je bilo mogoče dokaj natančno rekonstruirati nekdanjo naselbino, pri večini stavb pa celo ločiti število prostorov in razbrati njihov razpored (sl. 1.7). Mesto naselbine je bilo izbrano izredno premišljeno: večji del bivalnih stavb leži v nekoliko uleknjenem delu temena hriba, na vzvišenem skalnem platoju pa je bil že pred izkopavanjem dobro viden največji ruševinski sklop treh cerkva. obris četrte cerkve se nakazuje nekoliko nižje, prav tako ob povsem izpostavljenem jugozahodnem skalnem robu naselbine (glej pogl. 1.3). Velika poznoantična naselbina je ob odkritju pomenila precejšnje presenečenje, saj so bile dotlej tovrstne naselbine v Posočju slabo poznane. Arheološka podoba poznoantičnega obdobja je tu sicer dokaj pestra in nakazuje podobno kot v večjem delu Slovenije paleto naselbinskih tipov, vse od nezavarovanih naselij v 4. st. (glej pogl. 5.1) pa do močno utrjenih bivalnih središč na naravno zavarovanih položajih v 6. st. (glej pogl. 1.5, 5.2). Manjše naselbine iz 5. in 6. st. so bile odkrite v bližnji okolici Tonovcovega gradu in v njih je mogoče videti nekakšna satelitska naselja, ki so gravitirala prav k njemu. The archaeological site of Tonovcov grad, discovered in the upper part of the Soča Valley, was settled periodically from the Prehistory to the Middle Ages (see chapter 2.2). Its importance rose in the Late Antiquity when one of the largest and very complex fortified settlements of cross-regional importance developed here. The investigations of the hill have been going on since 1993, with short discontinuations (see chapter 1.4). They revealed a Late Antique settlement with approximately 30 buildings and a defence wall with towers on the more easily accessible parts (see chapter 1.3). On the northwest part of the hill there was also a large area, fortified by steep slopes and a defence wall, which could offer additional protection or serve as a refuge if necessary. The fortified settlement covers the top of an isolated rock pinnacle in the narrowest part of the Soča river gorge and measures 150 x 90 m. The 80 cm thick defence wall is especially well visible on the northeast side where it separates the settlement from the lower eastern part of the hill. Inside the settlement the ruins of buildings are visible in the surface relief. We were able to map the settlement with a certain degree of precision and in some cases it was even possible to discern the number of the individual spaces in a building and their layout (Fig. 1.7). The area where the settlement was built was carefully chosen: the majority of the dwelling houses lie in a small depression on the top of the hill, while the largest ruin of a church complex - well visible even before the excavations - is situated to the southeast on a high rocky plateau. The outline of a fourth church can be seen somewhat lower, but still by the completely exposed rocky edge of the settlement in the southwest (see chapter 1.3). The discovery of the large Late Antique settlement was a surprise as such settlements were poorly known in the Posočje region until then. However, the Late Antique settlement pattern in this region is quite varied and, similarly to a large part of Slovenia, shows a number of settlement types from unfortified settlements in the 4th century (see chapter 5.1) to the strongly fortified centres on naturally protected positions in the 6th century (see chapters 1.5, 5.2). Small settlements from the sistematična raziskava delno poškodovane velikega zgradbe (objekt št. 1) na Tonovcovem gradu na najugodnejšem, izravnanem in zatišnem delu naselbine blizu vhoda je pokazala obstoj dveh različnih objektov iz dveh poznoantičnih faz (fazi pA 1 in pA 2; glej pogl. 2.3). skromni ostanki zidov kažejo na obstoj zidane stavbe (ali celo dveh stavb) že v času prve intenzivnejše poselitve hriba, torej v poznorimskem obdobju (faza pA 1). oblika in velikost stavbe oziroma stavb nista znani, analiza najdb pa je pokazala, da jih je mogoče datirati v drugo polovico 4. in začetek 5. st. Najbolje časovno opredeljiva je uvožena keramika. Amfore in sigilata so prihajale iz afriškega prostora, glazirana keramika pa verjetno iz italskega (glej Tonovcov grad. najdbe, pogl. 4.1). Druge drobne najdbe so skromne in kažejo bivalni značaj. Značilni kosi vojaške opreme in fibule iz tega časa so bili najdeni večinoma pomešani v naselbinskih plasteh mlajše stavbe (glej pogl. 2.3.2 in Tonovcov grad. najdbe, pogl. 2.1). propada starejše stavbe ne moremo natančneje opredeliti, zadnje zanesljivo datirane najdbe sodijo v prva tri desetletja 5. st (glej pogl. 3.1 in Tonovcov grad. najdbe, pogl. 2.1.3). Mnogo jasnejša je slika, ki jo dajejo ostanki zidov stavbe 1 iz konca 5. in iz 6. st. V začetku je bila zgradba koncipirana kot enoprostorna velika zidana hiša z vetrolovom in nadstreškom, pozneje so ji dodali manjši, slabše izdelan stranski prostor. nastala je ob koncu 5. st. ali celo v začetku 6. st (glej pogl. 3.1.3). Glede na lego blizu glavnega vhoda bi lahko sklepali, da je bila njena funkcija povezana z nadzorom vhodnega območja. primerjava z drugimi sočasnimi zgradbami v naselbinah vzhodnoalpskega območja in Balkana kaže, da predstavlja eno izmed večjih in pomembnejših stanovanjskih zgradb, kar posredno potrjuje tudi njena lega znotraj naselbine (glej pogl. 4.1). Velika količina delov noše, orodja in hišne opreme dobro oriše podobo v veliki meri samooskrbne naselbine romanskih staroselcev, ki so se preživljali s poljedelstvom. Luksuzni predmeti in takšni, ki bi jih lahko pripisali Germanom, so redki. Večina najdb se odlično vklaplja v sliko ostalih znanih staroselskih naselbin v jugovzhodnih Alpah (glej Tonovcov grad. najdbe, pogl. 2.1). Med uvoženo keramiko prevladujejo vzhodnosredozemske amfore (LRA 1, LRA 2, LRA 4 in LRA 5/6), čeprav še obstajajo tudi vezi z afriškim prostorom. najmlajše gradivo predstavljata odlomek poznega malega spatejona in domnevne amfore tipa samos - cisterna. oba sta značilna za drugo polovico 6. in začetek 7. st. (glej Tonovcov grad. najdbe, pogl. 4.1). podobno kot keramika tudi steklene najdbe povezujejo naselbino z bizantinskim sredozemljem in nakazujejo na še obstoječe, vendar očitno drugače organizirane trgovske povezave in poti (glej Tonovcov grad. najdbe, pogl. 3.1). V nadaljevanju raziskav je bil odkrit stavbni kompleks treh zgodnjekrščanskih cerkva (glej pogl. 2.5, 3.3 in 4.2). Dve (severna in osrednja) sta bili povezani 5th and 6th century were discovered in the near vicinity of Tonovcov grad and can be interpreted as its satellites. A partly damaged large building (building 1) is situated on the most favourable, level part of the settlement which is shielded from the winds and near the entrance. systematic research showed the existence of two buildings from two late Antique phases (LA 1 and LA 2) in the excavation area of building 1 (see chapter 2.3). Modest wall remains suggest the existence of one (or even two) masonry buildings as early as during the first more intensive settlement of the hill, i.e. in the late Roman period (LA 1 phase). Their ground plan and size are not known. The analysis of the small finds (especially imported pottery) showed that the earlier building (or buildings) can be dated to the second half of the 4th and the beginning of the 5th century. Amphorae and African Red slip Ware were imported from Africa, while glazed pottery probably came from Italy (see Tonovcov grad. Finds, chapter 4.1). other small finds are modest and indicate that the building was used as a dwelling house. Typical parts of military equipment and fibulae of the Late Roman period were mostly found mixed in settlement layers of the later building 1 (see chapter 2.3.2 and Tonovcov grad. Finds, chapter 2.1). The time of abandonment of the earlier building is hard to define; the last datable finds belong to the first three decades of the 5th century (see chapter 3.1 and Tonovcov grad. Finds, chapter 2.1.3). The wall remains of building 1 from the end of the 5th and the 6th century give a much clearer picture. The building was first built as a large single-space masonry house with a wind shield and a porch. Later a small outhouse of poorer masonry was added. Building 1 was built at the end of the 5th or even at the beginning of the 6th century (see chapter 3.1.3). As it is situated close to the main entrance to the settlement it could have been used to help monitor the entrance area. Comparisons with other contemporary buildings in the settlements of the southeastern Alps and the Balkans show that building 1 belongs to the larger and more important dwelling houses, which is also indirectly confirmed by its position within the settlement (see chapter 4.1). The large quantity of objects of attire, tools and architectural fittings, found in the building, reflects the largely autarkic settlement of the Roman autochthonous population whose main source of livelihood was agriculture. luxury objects, or artefacts that could be ascribed to the Germanic peoples, are rare. Most small finds fit very well into the picture formed by the other known settlements of the autochthonous population in the southeastern Alps (see Tonovcov grad. Finds, chapter 2.1). Among the imported pottery eastern Mediterranean amphorae are predominant (LRA 1, LRA 2, LRA 4 and LRA 5/6), but links with Africa are still existent (see Tonovcov grad. Finds, chapter 4.1). The latest pieces are a fragment of a small late spatheion and an amphora neposredno, tretjo (južno), ki je bila zgrajena kasneje, pa je z osrednjo povezoval pravokoten zidan prostor med prezbiterijema. Vse tri cerkve na Tonovcovem gradu imajo preprosto pravokotno zasnovo s prosto stoječo klopjo za duhovnike in so po svoji oblikovanosti tipičen predstavnik akvilejskega kroga. V cerkvah na Tonovcovem gradu so bili vsi trije oltarji zidani, kar je posebnost v vzhodnoalpskem prostoru. oltarji in dobro ohranjene klopi za duhovnike so bili dodani ob prenovi sredi 6. st., ki je dobro datirana z zakopom spatejona z Justinijanovim novcem v razpoki ob oltarju osrednje cerkve. Velik problem predstavlja identifikacija baptisterija oziroma njegove piscine v cerkvenem sklopu. V začetni fazi raziskav je bila piscina domnevana v umetno razširjeni skalni razpoki pred prezbiterijem osrednje cerkve predvsem zaradi ustrezne velikosti jame in pa domnevno pripadajoče zravnane površine v prezbiteriju tik nad njo. Danes se zdi domneva verjetna predvsem za zgodnejšo fazo obstoja cerkva, poizkusi locirati baptisterij v druge prostore pa niso zanesljivi. Še vedno obstaja tudi manjša možnost, da leži baptisterij nekoliko dlje stran od cerkva, v še neizkopanem predelu (objekt št. 29, glej sl. 1.7). Pomena tudi ni mogoče zanesljivo določiti za pravokoten prostor, ki je bil na južni strani prizidan na osrednjo cerkev. Glede na okrašen vhod in kenotaf je mogoče sklepati, da gre za memorijo. Grobovi v cerkvi in ob njej so značilni za zgodnje-krščansko miselnost in izražajo željo biti pokopan ad sanctos. En grob je bil najden v prezbiteriju južne cerkve, eden v ladji osrednje, dva pa v narteksu severne in osrednje cerkve. Ob narteksu južne cerkve je ležal ženski grob z elementi noše, ob vzhodnih stranicah severne in osrednje cerkve ter memorije pa je bilo v skalo vklesanih še sedem otroških grobov brez pridatkov. Raziskave na Tonovcovem gradu so dale kar nekaj bolje datiranih elementov, ki v grobem določajo časovni okvir cerkva od konca 5. do začetka 7. st. in predvsem pomembno dodelavo sredi 6. st. (glej pogl. 3.3). Tako datacijo posredno potrjuje tudi astronomska orientacija južne (zadnje zgrajene) cerkve proti Sončevemu vzhodu na 2. februar julijanskega koledarja v 6. st. (glej pogl. 4.3). To je praznik svečnice, katerega čaščenje je bilo v bizantinskem imperiju uradno uvedeno po letu 542. Namembnost cerkvenega sklopa na Tonovcovem gradu je težko dokončno opredeliti, kajti pripadajoči bivalni in eventualni reprezentančni prostori, ki se nakazujejo v bližini, še niso raziskani. Tri cerkve, kakovostna gradnja in različne liturgične funkcije pa ob tem, da daleč naokoli ni podobnega zgodnjekrščanskega sklopa, dovoljujejo sklep o pomembnem cerkvenem regionalnem središču (glej pogl. 4.2) V bližini cerkva je bila izkopana še ena stavba (objekt št. 2), ki je za svojo severno stranico izkoristila steno starejše zelo uničene stavbe (objekt št. 3; glej pogl. 2.4). Stavba 2 je bila brez ognjišča, tudi najdbe so izredno fragment of presumably the Samos-cistern type. Both are typical of the second half of the 6th and the beginning of the 7th century (see Tonovcov grad. Finds, chapter 4.1). Similar to pottery, glass finds link the settlement with the Byzantine Mediterranean and point to the still existing, but apparently differently organized, trade connections and routes (see Tonovcov grad. Finds, chapter 3.1). With the continuation of research a building complex of three Early Christian churches was discovered (see chapters 2.5, 3.3 and 4.2). Two (the north and main) churches were joined in their entire length, while the third (south church), which was built later, was linked with the main church by a square stone structure between the two presbyteries. All three churches on Tonovcov grad have a simple rectangular single-nave layout with a free standing bank for the clergy; their ground plan is typical of the Aquileia circle. In the churches on Tonovcov grad all three altars were masonry structures, which is rare in the Southeastern Alps. The altars and the well preserved banks for the clergy were added during a renovation phase in the middle of the 6th century, which is dated with a spatheion with a coin of Justinian, buried in a rock crevice near the altar of the main church. The identification of the baptistery or its piscina in the church complex is a problem. In the early phase of the research we assumed that the piscina was in the artificially widened rock crevice in front of the presbytery of the main church. This was mostly because the size of the opening seemed appropriate and because there is a small levelling in the rock floor just above the opening which could belong to the baptistery. Today this assumption seems probable especially for the early phase of the complex; the attempts at locating the baptistery in the other rooms were not convincing. There is still a possibility that the baptistery was located a little further away from the churches in the unexplored area (building No. 29, see Fig. 1.7). It was not possible to define the function of the lateral square space which was added to the main church on the south side. Judging by the decorated entrance and a cenotaph it could be interpreted as a memoria. Graves inside and outside a church are typical of the Early Christian mentality and express the wish to be buried ad sanctos. One grave was found in the presbytery of the south church, one in the nave of the main church and two in the narthex of the north and main churches. A female grave with jewellery lay by the narthex of the south church and seven children's graves without grave goods were cut into the rock by the east walls of the north and main church and the 'memoria'. The research on Tonovcov grad offered quite a few well dated elements which define the time frame of the churches roughly between the end of the 5th and the beginning of the 7th century, and above all the important renovation in the middle of the 6th century skromne, tako da verjetno ni bila stanovanjski objekt. Nekateri časovno bolje opredeljivi predmeti iz njene okolice jo datirajo v 6. st., in sicer bolj v njegovo drugo polovico. starejša stavba 3 sodi v prvo poznoantično fazo (druga polovica 4. in začetek 5. st.), pozneje (v drugi poznoantični fazi) pa je bila njena ruševina uporabljena pri gradnji stavbe 2. Na vzhodnem platoju naselbine, ki se še nekoliko dviga nad plato s cerkvami, je bila raziskana velika vodna cisterna (glej pogl. 2.6, 3.4). Zaradi skromnih najdb je začetek njene izgradnje težko natančno določiti, najverjetneje je bila zgrajena hkrati s cerkvenim sklopom, torej konec 5. st. V naselbini odkrito drobno gradivo, novci in arhitektura omogočajo zanesljiv kronološki okvir (glej pogl. 2.3), ki nakazuje prvo poselitev že v prazgodovini (glej Tonovcov grad. Najdbe, pogl. 6) prva uporaba hriba v antičnem obdobju se nakazuje v zadnji tretjini 3. st., podobno, kot je bilo to mogoče zaslediti tudi na številnih drugih najdiščih v sloveniji, pa tudi v bližnji Furlaniji in na Koroškem. Vendar pa iz tega časa doslej nismo našli ostankov bivalne arhitekture, zato je mogoče sklepati, da gre za kratkotrajno zadrževanje prebivalstva na že po naravi dobro zavarovanem kraju. prva pomembnejša sklenjena poselitev na Tonov-covem gradu je nastopila v drugi polovici 4. in v prvih desetletjih 5. st. (faza pA 1). o tem pričajo skromni arhitekturni ostanki ob stavbi 1, ostanek stavbe 3, izpovedni kovinski predmeti, novci in keramika. Izredno številne in bogate elemente moške noše ter ostanke arhitekture iz te faze na Tonovcovem gradu je mogoče razložiti le s prisotnostjo vojaške posadke. Nanjo kažejo tudi dokaj številne amfore iz druge polovice 4. st., v katerih so iz Afrike v velika pristanišča, od tam pa tudi v zaledje, dostavljali živila za vojsko (glej Tonovcov grad. najdbe, pogl. 4.1). Tonovcov grad leži na križišču pomembnih poti med italijo in zaledjem. kratek pregled cestnih povezav in utrdb ob njih v severozahodni sloveniji je pokazal, da se je življenje v njih intenziviralo že v 4. st. in še posebej v času, ko je bila opuščena glavna magistrala z vzhoda preko Hrušice v italijo (glej pogl. 5.1). Takrat so utrdbe ob sekundarnih komunikacijah morale prevzeti breme pritiskov, ki jih je prej v veliki meri odbijal tristopenjski sistem obrambnih zidov in trdnjav med Vrhniko in Hrušico. Vojaško utrdbo na Tonovcovem gradu je zato mogoče razumeti le v sklopu zapornega sistema Claustra oziroma v času funkcioniranja vojne cone, poimenovane tractus Italiae circa Alpes. V poznorimskih pisnih virih se omenjajo tako claustra, torej umetni zaporni zidovi, kot tudi clausurae, kar večina zgodovinarjev tolmači kot naravno težko prehodna mesta (soteske, ožine, prehodi), ki jih je bilo mogoče uspešno zavarovati. Arheološki viri kažejo, da je bil poudarek obrambe na umetnih linearnih zidovih predvsem v celotnem 4. st., medtem (see chapter 3.3). This date is indirectly confirmed by the astronomical orientation of the south church (the last to be erected) towards the rising of the sun on February 2nd of the Julian Calendar in the 6th century. This is the day of Candlemas which was officially celebrated in the Byzantine Empire after 542 (see chapter 4.3). The function of the church complex on Tonovcov grad is difficult to define precisely as the dwelling and potential representative buildings, which are outlined in the surface relief in the vicinity, are not yet researched. The three churches, good quality masonry and different liturgical functions, together with the fact that there is no comparable early Christian complex in the vicinity, allow for the conclusion that this was an important regional ecclesiastical centre (see chapter 4.2). Near the churches another building (building 2) was researched. As its north wall this building used the wall of an earlier, heavily destroyed building (building 3; see chapter 2.4). Building 2 had no fireplace and the small finds are very modest, so we assume it was not used as a dwelling house. It is dated to the 6th century, even to its second half, by some better definable objects. The earlier building 3 belongs to the first late Antique phase (phase lA 1: second half of the 4th-beginning of the 5th century), its ruin was later used to construct building 2. on the eastern plateau of the settlement, which rises above the church plateau, a large water cistern was investigated (see chapters 2.6, 3.4). it was probably built at the same time as the ecclesiastical complex at the end of the 5th century but because of the small number of finds this is difficult to confirm. The small finds, coins and architecture, discovered in the settlement, allow for a reliable chronological frame (see chapter 2.3) which indicates the first settlement of the hill already in the prehistory (Tonovcov grad. Finds, chapter 6). The first occupation of the hill in Antiquity is indicated in the last third of the 3rd century, similarly as on many other sites in slovenia and also in Friuli and Carinthia. As no architecture belonging to this period has so far been found, we must presume that this occupation was only a short stay of the population on the naturally well defended place. The first more important settlement of Tonovcov grad is dated to the second half of the 4th and the first decades of the 5th century (phase lA 1). The remains of this phase are the walls in the excavation area of building 1, remains of building 3 and some metal finds, coins and pottery. numerous and rich elements of male attire and architectural remains of this phase on Tonovcov grad can only be explained by the presence of a military garrison. This idea is confirmed by numerous amphorae from the second half of the 4th century which were used to transport military provisions from Africa to large ports and from there on to the hinterland (see Tonovcov grad. Finds, chapter 4.1). ko se v 5. st. omenjajo večidel clausure. To se dobro ujema s sliko, ki jo kažejo predvsem nekatere utrdbe ob pomembnejših komunikacijah: najbolje raziskani Tonovcov grad, puštal pri Trnju in Ajdovski gradec pri Bohinjski Bistrici. Glede na številne odkrite predmete vojaške noše in strateški položaj bi smeli vsaj pri prvih dveh domnevati postojanke treh legij, znanih iz Noticije Dignitatum, poimenovanih Iuliae Alpinae. poznorimske najdbe na širšem območju Kobarida kažejo na močno povečano vlogo prometnega vozlišča pred enim izmed najlažjih prehodov iz doline soče v Furlanijo. Dovoljujejo domnevo, da je nastal na tem mestu sklop naselbin in utrdb, ki so imele nalogo nadzirati premike skozi ta prostor in po potrebi še zadnjič zadržati sovražnika pred vrati Italije. pri tem pa je naloga najpomembnejšega branika ceste, ki je tekla čez predel, pripadla prav utrdbi na Tonovcovem gradu. posredno pomembnost ceste in naselbine potrjuje močno povečan pomen bližnjega Čedada v poznoantičnem času, še posebej po prihodu Langobardov. utrjena naselbina na Tonovcovem gradu iz druge poznoantične faze z zidano stavbno arhitekturo, obzidjem s stolpi in zgodnjekrščanskim središčem na vzvišenem mestu je izrazit in pomemben predstavnik postojank sklepne faze pozne antike na vzhodnoalp-skem območju, hkrati pa je tudi ena izmed najbolje ohranjenih. V sočasni poselitveni sliki izstopajo trije aspekti, ki določajo njen pomen in vlogo: premišljeno koncipirana obsežna utrjena naselbina, izjemna strateška lega in prisotnost sakralnega središča (glej pogl. 5.2). kažejo, da gre za eno izmed redkih postojank, v kateri so združene značilnosti različnih tipov pozno-antičnih višinskih utrdb. predstavljene primerjave s sočasnimi sorodnimi najdišči na vzhodnoalpskem in zahodnobalkanskem območju kažejo, da gre na Tonovcovem gradu za eno izmed najbolj kompleksnih in pomembnih središč zadnjega obdobja antike na vzhodnoalpskem območju. Zaradi svoje odlične naravne obrambne lege kot tudi strateške umeščenosti je bila postojanka primerna za osrednjo naselbino Zgornjega posočja in hkrati je omogočala posadki v njej nadzor prometa po pomembni vpadnici proti italiji. novci in posamezni predmeti v primerjavi z drugimi analognimi utrdbami nakazujejo občasno prisotnost manjših germanskih posadk v naselbini, predvsem vzhodnogotske in langobardske. V naselbini so postopno dograjevali tudi celovit sklop cerkva, ki predstavlja sakralno središče širše regije in ima delno tudi romarski značaj. Analiza naselbin z dvojnimi cerkvami v višinskih utrjenih naselbinah vzhodnoalpskega prostora skupaj s specifičnimi pripadajočimi bivalnimi in reprezentančnimi stavbami je opozorila na izjemnost tovrstnih najdišč v poznoantični naselbinski sliki (glej pogl. 5.2). naštetim elementom je mogoče dodati še funkcijo pribežališča, kar arheološko v naselbini sicer težko potrdimo, posredno pa jo nakazuje nepozidan utrjen prostor pred njo. A short review of road connections and fortifications on these roads in northwest slovenia shows that a more intensive occupation began already in the 4th century, and especially in the time after the abandonment of the main road from the east across Hrušica (Ad pirum) to Italy (see chapter 5.1). Then the fortifications on secondary roads had to take the burden of the pressures which had earlier been largely repelled by the three-phase system of the defence walls and fortifications between Vrhnika (Nauportus) and Hrušica (Ad pirum). The fort on Tonovcov grad can therefore only be understood within the Claustra defence system or in the time of the war-zone called Tractus Italiae circa Alpes. In the Late Roman written sources, mention is made of claustra, artificial defence walls, as well as clausurae, interpreted by most historians as places which are naturally difficult to pass (gorges, straits, and passes), but can be successfully defended. Archaeological sources show that emphasis of the defence on artificial linear walls is noticeable especially in the 4th century, while in the 5th century clausurae are most often mentioned. This seems to be confirmed by some fortifications on important communications: (the best researched) Tonovcov grad, puštal near Trnje and Ajdovski gradec near Bohinjska Bistrica. Regarding the numerous objects of military attire and the strategic positions of the forts it could be assumed, at least for the first two, that they were posts of the three legions called Iuliae Alpinae, known from the Notitia dignitatum. Late Roman finds in the wider territory of Kobarid show the increased importance of the crossroads in front of one of the easiest passes from the soča Valley to the Friuli. They allow for the assumption that a complex of settlements and forts was created here which monitored the movements through this territory and, if necessary, functioned as the last defence line before Italy. The most important fort on the road across the predel/predil pass was Tonovcov grad. The importance of the road and the settlement is indirectly confirmed by the increased importance of the nearby Cividale del Friuli in the Late Antiquity, especially after the arrival of the Lombards. The fortified settlement of Tonovcov grad in the second Late Antique phase with its masonry architecture, defence wall with towers and an Early Christian centre on an elevated position, is an important representative of the last Late Antique phase fortifications in the southeastern Alps, while at the same time it is one of the best preserved (see chapter 5.2). Three aspects, which define its importance and role in the contemporary settlement pattern, must be mentioned: a deliberately conceived large fortified post, exceptional strategic position and the presence of an ecclesiastical centre. These aspects show that this is one of the rare sites where characteristics of different types of Late Antique fortified hilltop posts are joined together. The presented analogies with similar contemporary sites in V obeh obdobjih pozne antike je bil Tonovcov grad ekonomsko avtarkična enota, katere obstoj je v veliki meri temeljil na subsistenčnem gospodarstvu (govedoreja, reja drobnice, prašičereja, reja kokoši; glej Tonovcov grad. Najdbe, pogl. 8). Antropološka analiza človeških skeletov zaradi majhnega vzorca ni reprezentativna, pokazala pa je med drugim tudi izredno nizko frekvenco kariesa (glej Tonovcov grad. Najdbe, pogl. 7). utrjena naselbina na Tonovcovem gradu pri Kobaridu predstavlja eno izmed naselbinskih žarišč, ki ji je v veliki meri uspelo zadržati prvine antične civilizacije z znaki občasne prisotnosti germanskih elementov vse do začetka 7. st. V spremenjeni naselbinski strukturi 6. st. ji je tako mogoče pripisati vlogo, kakršno so imela transformirana poznoantična mesta. Naselbina je bila verjetno opuščena postopno v prvi polovici 7. st., saj ni bila ugotovljena strnjena žganinska plast, ki bi kazala na enkratno katastrofo. To potrjuje tudi pomanjkanje dragocenejše opreme v cerkvah, posebej relikvijarijev. Nekaj najdb nakazuje krajše zadrževanje ljudi med ruševinami naselbine v zgodnjem srednjem veku. Zaradi skromnih ostankov lahko te obiske postavimo le okvirno med 7. in 9. st. Zadnji znak človeške prisotnosti je nekaj posamičnih najdb, ki spadajo v visoko- in poznosred-njeveško obdobje. the Southeastern Alps and the western Balkans show that Tonovcov grad is one of the most complex and important centres of the last phase of Late Antiquity in the Southeastern Alps. Because of its excellent naturally protected position and strategic location, the post was appropriate for the central settlement of the Upper Soča Valley and at the same time it allowed the garrison, which was stationed in it, to control the important incursion road towards Italy. Some coins and individual finds, when compared to analogous forts elsewhere, point to a temporary presence of small Germanic garrisons in the settlement, especially the Eastern Goths and Lombards. The ecclesiastical complex, which represents the centre of a wider region and has in part also the characteristics of a pilgrimage site, was gradually constructed. The analysis of settlements with double churches in the fortified hilltop settlements in the Southeastern Alps, together with specific dwelling and representative buildings showed that such sites are exceptional in the Late Antique settlement pattern (see chapter 5.2). To the listed elements we can add the function of a refuge which is difficult to confirm archaeologically, but it is implied by an empty fortified space in front of the settlement. In both periods of the Late Antiquity, Tonovcov grad was an economically autarkic unit which existed on subsistence economy (cattle, sheep, goats, pigs and poultry; see Tonovcov grad. Finds, chapter 8). Anthropological analysis of the human skeletons is not representative due to the small sample, but it did show a very low caries frequency (see Tonovcov grad. Finds, chapter 7). The fortified settlement of Tonovcov grad represents one of the settlement foci which was capable of retaining - to a large degree - the characteristics of the Antique civilisation with signs of temporary presences of Germanic elements as long as until the beginning of the 7th century. In the changed settlement structure of the 6th century it can thus be ascribed the role of the transformed Late Antique towns. The settlement was probably gradually abandoned in the first half of the 7th century, as no certain signs of a large fire, which would indicate a one-time catastrophe, were found. This is also confirmed by the lack of more valuable finds, especially reliquaries in the churches. Some finds indicate short visits of people among the ruins of the settlement in the Early Middle Ages. Because of the scarcity of finds these visits can only be placed to the time between the 7th and the 9th century. The last signs of human presence are some individual finds that belong to the High- and Post-medieval period. 7. literatura 7. bibliography OKRAJŠAVE / ABBREVIATIONS Tonovcov grad. Najdbe / Tonovcov grad. Finds = MODRIJAN, Z. in /and T. MILAVEC 2011, Poznoantična utrjena naselbina Tonovcov grad pri Kobaridu. Najdbe / Late Antique fortified settlement Tonovcov grad near Kobarid. Finds. - Opera Instituti Archaeologici Sloveniae 24. ZRC SAZU: Znanstvenoraziskovalni center Slovenske akademije znanosti in umetnosti / Scientific Research Centre of the Slovenian Academy of Sciences and Arts ZVNKD: Zavod za varstvo naravne in kulturne dediščine Slovenije/ Institute for the Protection of Natural and Cultural Heritage of Slovenia ABRAHAMSEN, N. 1992, Evidence for church orientation by magnetic compass in twelfth-century Denmark. - Archaeometry 34, 293-303. ALZINGER, W. 1977, Das Municipium Claudium Aguntum. Vom keltischen Oppidum zum frühchristlichen Bischofssitz. - V / In: Aufstieg und Niedergang der römischen Welt II/6, 380-413. ANDJELIC, T. 1978, Kasnoantičke bazilike u Cimu i Žitomisličih kot Mostara. - Arheološki vestnik 29, 629-640. ANSl 1975, Arheološka najdišča Slovenije. - Ljubljana. ATLAS SLOVENIJE 1986. - Ljubljana. AVENI, A. F. 1981, Archaeoastronomy. - V / In: M. B. Schiffer (ur./ ed.) Advances in archaeological method and theory 4, 1-77, New York. AVENI, A. F. 2001, Skywatchers: A revised and updated version of Skywatchers of ancient Mexico. - Austin. AVENI, A. F. 2003, Archaeoastronomy in the ancient Americas. - Journal of Archaeological Research 11 (2), 149-191. BARLAI, K. 1989, Some archaeoastronomical problems in East-Central Europe. - V / In: A. F. Aveni (ur. / ed.), World archaeoastronomy, 436-440, Cambridge. BASLER, D. 1972, Arhitektura kasnoantičkog doba u Bosni i Hercegovini. - Sarajevo. BASLER, D. 1993, Spätantike und Frühchristliche Architektur in Bosnien und der Herzegowina. - Schriften der Balkankommission, Antiquarische Abteilung 19. BAUSOVAC, M. 2003, Poznoantični kovinski predmeti z Gradca pri Prapretnem. - Arheološki vestnik 54, 315-324. BECHTEL, F. 1910, Itineraria. - V / In: The Catholic Encyclopedia, New York (http://www.newadvent.org/ cathen/08254a.htm [zadnji dostop / last checked 8. 10. 2011]). BELMONTE AVILES, J. A. in / and M. HOSKIN 2002, Reflejo del cosmos: Atlas de arqueoastronomia en el Mediterräneo antiguo. - Madrid. BENSON, H. 1956, Church orientations and patronal festivals. - The Antiquaries Journal 36 (3-4), 205-213, London. BIERBRAUER, V. 1987, Invillino-Ibligo in Friaul 1, Die römische Siedlung und das spätantik-frühmittelalterliche Castrum. - Münchner Beiträge zur Vor- und Frühgeschichte 33. BIERBRAUER, V. 1988, Invillino-Ibligo in Friaul 2, Die spätantiken und frühmittelalterlichen Kirchen. -Münchner Beiträge zur Vor- und Frühgeschichte 34. BIERBRAUER, V. 1998, Arianische Kirchen in Noricum mediterraneum und Raetia II? - Bayerische Vorgeschichtsblätter 63, 205-226. BOLTA, L. 1981, Rifnik pri Šentjurju. Poznoantična naselbina in grobišče. - Katalogi in monografije 19. BONA, I. 1970-1971, Langobarden in Ungarn. -Arheološki vestnik 21-22, 45-74. BONETTO, J. in / and L. VILLA 2003, Nuove con-siderazioni sulle cinte fortificate di Forum Iulii alla luce dello scavo di casa Canussio. - Forum Iulii 27, 15-67. BOSIO, L. 1957, Due tratti di strada romana sco-perti nella X regio. - Aquileia Nostra 26-28, 29-38. BOSIO, L. 1970, Itinerari e strade della Venetia romana. - Padova. BOSIO, L. 1991, Le strade romane della Venetia e dell'Histria. - Padova. BOŽIČ, D. in I and S. CIGLENEČKI 1995, Ze-nonov tremis in poznoantična utrdba Gradec pri Veliki Strmici. - Arheološki vestnik 46, 247-277. BOŽIČ, Z. 1993, Novi kobariški čudež. - Primorske novice, 26. 3.1993, priloga 7. Val. BRANK, R. 1979, Sledovi nasipov poznoantičnih obrambnih zapor pri Novi Oselici. - Loški razgledi 26, 18-23. BRATINA, P. 1999, Golo Brdo. - Varstvo spomenikov 38 - Poročila (izšlo 2001), 28-30. BRATOŽ, R. 1991, Krščanstvo v antiki. - V I In: M. Benedik (ur.I ed.), Zgodovina Cerkve na Slovenskem, 11-27, Celje. BRATOŽ, R. 1994, Bitka pri Frigidu v izročilu antičnih in srednjeveških avtorjev. - Zbirka Zgodovinskega časopisa 12, 5-45, Ljubljana. BRATOŽ, R. 1996, Doppelkirchen auf dem östlichen Einflußgebiet der aquileiensischen Kirche und die Frage des Einflusses Aquileias. - Antiquité tardive 4, 133-141. BRATOŽ, R. 2001, Soča in prehodi čez reko v antiki. - Goriški letik 27 (2000), 27-50. BRATOŽ, R. 2003, Aquileia tra Teodosio e i Lon-gobardi (397-568). - V I In: Aquileia dalle origini alla costituzione del ducato Longobardo (Atti della XXXIII settimana di Studi Aquileiesi, 25-27 aprile 2002). Societa Antichità Altoadriatiche 54, 477-521. BROGIOLO G. P. 2001, Gli scavi a Monte Barro 1990-97. - V I In: G. P. Brogiolo, L. Castelletti. (ur. I eds.), Archeologia a Monte Barro. II - Gli scavi 1990-97 e le ricerche al S. Martino di Lecco, 21-102. BROGIOLO, G. P., UBOLDI, M. in I and G. BERNARDI 2001, Lo scavo dell'insediamento rupestre del Monte S. Martino di Lecco. - V I In: G. P. Brogiolo, L. Castelletti (ur. I eds.), Archeologia a Monte Barro. II -Gli scavi 1990-97 e le ricerche al S. Martino di Lecco, 287-307. BUSER, S., RAMOVŠ, A., DROVENIK, M. in I and M. PLENIČAR 1995, Geološki razvoj Slovenije. - V I In: Enciklopedija Slovenije, 3. zvezek, 195-203, Ljubljana. CAGNANA, A. 2000, Gli scavi nel complesso cultuale di San Martino di Ovaro (Carnia). Notizia preliminare. - Forum Iulii 24, 23-28. CAGNANA, A. 2008, Testimonianze della cri-stianizzazione in Carnia. - V I In: S. Piussi (ur. I ed.), Cromazio di Aquileia 388-408: al crocevia di genti e religioni, 440-447, Milano. CAMBI, N. 2005, Srima i dvojne bazilike u Dalmaciji. - V I In: Z. Gunjača et al., Srima - Prižba. Starokrščanske dvojne cerkve, 57-70, Šibenik. CANTINO WATAGHIN, G. 1996, Le "basiliche doppie" paleocristiane nell'Italia settentrionale: la docu-mentazione archeologica. - V I In: Antiquité tardive 4, 115-123. CAVE, C. J. P. 1950, The orientation of churches. -The Antiquaries Journal 30 (1-2), 47-51. CEDILNIK, A. 2004, Ilirik med Konstantinom Velikim in Teodozijem Velikim. - Thesaurus memoriae. Dissertationes 3. CHEVALIER, P. 1995, SalonaII. EcclesiaeDalmati-ae. - Collection de l'Ecole française de Rome 1994, vol. 1, 2, Roma, Split. CHRISTIE, N. 1991, The Alps as a frontier (A. D. 168-774). - Journal of Roman Archaeology 4, 410-430. CHRISTIE, N. 1995, The Lombards. The Ancient Longobards. - OxfordICambridge. CIGLENEČKI, S 1975, Kasnoantično utrjeno naselje Gradec pri Prapretnem. - Arheološki vestnik 26, 259-267. CIGLENEČKI, S. 1979, Kastel, utrjeno naselje ali refugij? - Arheološki vestnik 30, 459-472. CIGLENEČKI, S. 1981, Rezultati prvih raziskovalna Gradcu pri Prapretnem. - Arheološki vesnik 32, 417-453. CIGLENEČKI, S. 1985, Potek alternativne ceste Siscija-Akvileja na prostoru zahodne Dolenjske in Notranjske v času 4. do 6. stoletja. Preliminarno poročilo o raziskovanjih Korinjskega hriba in rekognosciranjih zahodne Dolenjske. - Arheološki vestnik 36, 255-284. CIGLENEČKI, S. 1986, Zgodnjekrščansko središče na Kučarju pri Beli krajini. - Izdanja Hrvatskog areološkog društva 10, 137-150. CIGLENEČKI, S. 198ya, Höhenbefestigungen aus der Zeit vom 3. bis 6. Jh. im Ostalpenraum. - Dela 1. razreda SAZU 31. CIGLENEČKI, S. 198yb, Das Weiterleben der Spätantike bis zum Auftauchen der Slawen in Slowenien. - V I In: Die Völker Südosteuropas im 6. bis 8. Jahrhundert, Südosteuropa - Jahrbuch 17, 265-286. CIGLENEČKI, S. 1990, K problemu datacije nastanka rimskodobnih višinskih utrdb v jugovzhod-noalpskem prostoru. - Arheološki vesnik 41, 147-176. CIGLENEČKI, S. 1992a, Tracce di un insedia-mento tardo (VI-IX sec.) nei siti della tarda antichità in Slovenija. - Il territorio tra tardoantico e altomedioevo. Metodi di indagine e risultati, 53-59, Firenze. CIGLENEČKI, S. 1992b, Polis Norikon, Poznoantične višinske utrdbe med Celjem in Brežicami. - Podsreda. CIGLENEČKI, S. 1993, Še o novem "čudežu" v Kobaridu. Tonovcov grad sodi med najimenitnejša slovenska poznoantična najdišča. - Dom, leto XXVIII, št. 10, 31. maja 1993. CIGLENEČKI, S. l994a, Höhenbefestigungen als Siedlungsgrundeinheit der Spätantike in Slowenien. -Arheološki vestnik 45, 239-266. CIGLENEČKI, S. l994b, Poznoantična naselbina Tonovcov grad pri Kobaridu. - Kronika 42, 1-14. CIGLENEČKI, S. l994c, Scavi nell'abitato tardo-antico di Tonovcov grad presso Caporetto (Kobarid) Slovenia. Rapporto preliminare. - Aquileia Nostra 65, 185-208. CIGLENEČKI, S. 1995, Izvrednotenje / Auswertung. - V / In: J. Dular, S. Ciglenečki, A. Dular 1995, 134-190. CIGLENEČKI, S. 1997a, Die wichtigsten Völkerwanderungszeitlichen Einfallstrassen von Osten nach Italien im Licht der neuesten Forschungen. - V / In: Peregrinatio Gothica, Jantarova stezka, Supplementum ad Acta Musei Moraviae, Scientiae sociales 82, 179-191. CIGLENEČKI, S. 1997b, Poznoantični in slovanski sledovi na Kobariškem. - V / In: Kobarid, 17-28, Kobarid. CIGLENEČKI, S. 1997c, Strukturiranost poznorimske poselitve Slovenije. - Arheološki vestnik 48, 191-202. CIGLENEČKI, S. 1997d, Tonovcov grad pri Kobaridu. Arheološko najdišče. Vodnik. - Ljubljana - Kobarid. CIGLENEČKI, S. 1999, Results and Problems in the Archaeology of the Late Antiquity in Slovenia / Izsledki in problemi poznoantične arheologije v Sloveniji. - Arheološki vestnik 50, 287-309. CIGLENEČKI, S. 2000, Tinje nad Loko pri Žusmu. Poznoantična in zgodnjesrednjeveška naselbina / Tinje oberhalb von Loka pri Žusmu. Spätantike und frühmittelalterliche Siedlung. - Opera Instituti Archaeologici Sloveniae 4. CIGLENEČKI, S. 2001, Romani e Longobardi in Slovenia nel VI secolo. - V / In: Paolo Diacono e il Friuli altomedievale (secc. VI-X), 179-200, Spoleto. CIGLENEČKI, S. 2003, Frühchristliche Kirchen in Slowenien und die Elemente ihrer Innenausstattung. -Hortus artium medievalium 9, 11-20. CIGLENEČKI, S. 2005, Langobardische Präsenz im Südostalpen im Lichte neuer Forschungen. - V / In: W. Pohl, P. Erhart (ur. / eds.), Die Langobarden: Herrschaft und Identität. Forschungen zur Geschichte des Mittelalters, Bd. 9, 265-280. CIGLENEČKI, S. 2006, Zur Chronologie frühchristlicher Gebäude in Slowenien. - V / In: R. HARREITHER et al. (ur. / eds.), Frühes Christentum zwischen Rom und Konstantinopel. Acta congressus internationalis XIV archaeologiae christianae, Vindobonae 19.-26. 9. 1999, 293 -300. CIGLENEČKI, S. 2007, Zum Problem spätrömischer militärischer Befestigungen im südlichen Teil von Noricum Mediterraneum. - V / In: Schild von Steier 20, 317-328. CIGLENEČKI, S. 2008, Castra und Höhensiedlungen vom 3. bis 6. Jahrhundert in Slowenien. - V / In: H. Steuer, V Bierbrauer, M. Hoeper (ur. / eds.), Höhensiedlungen zwischen Antike und Mittelalter von den Ardennen bis zur Adria, 481-532, Berlin, New York. CIGLENEČKI, S. in /and T. MILAVEC 2009, The defence of north-eastern Italy in the first decennia of the 5th century. - Forum Iulii 33, 175-187. CIPOLLONE, V: 2006, Ricerche archeologiche degli anni '80 presso Nimis (UD). - Archeologia Medievale 33, 131-142. CONOLLY, J. in /and M. LAKE 2006, Geographical Information Systems in Archaeology. - Cambridge. CURK, I. 1994/1995, Za razvoj delovnih metod -arheološka dediščina. - Varstvo spomenikov 35/93, 76-77. CUSCITO, G. 2001, "In Castra gradensi ac plebe sua": Lo sviluppo del castrum di Grado dalla tarda antichita all'alto medioevo. - V / In: Paolo Diacono e il Friuli altomedievale (secc. VI-X), 387- 406, Spoleto. CVITKOVIČ, J. 1999, Arheološka topografija srednje in zgornje soške doline. - Diplomska naloga / Diploma. Oddelek za arheologijo, Filozofska fakulteta Univerze v Ljubljani, Ljubljana. ČAVAL, S. s.a., Astronomska usmerjenost romanskih cerkva v Sloveniji. - Neobjavljena doktorska disertacija / Unpublished Doctoral Dissertation. Oddelek za arheologijo, Filozofska fakulteta Univerze v Ljubljani, Ljubljana. ČUK, S. 1999, Svetnik za vsak dan. Januar - junij. - Koper. DE COULANGES, N. D. F. 1873 (2001), The Ancient City. A Study on the Religion, Laws, and Institutions of Greece and Rome, Baltimore. DEMANDT, A. 1989, Die Spätantike. Römische geschichte von Diocletian bis Justinian. - Handbuch der Altertumswissenschaft III.6, München. DOLENC, J. 1992, Zlati Bogatin. Tolminske povedke. - Ljubljana. DAKSKOBLER, I. 1995, Posočje. - V / In: Enciklopedija Slovenije, 9. zvezek, 167, Ljubljana. DULAR, J., KRIŽ, B., SVOLJŠAK, D. in / and S. TECCO HVALA 1991, Utrjena prazgodovinska naselja v Mirenski in Temeniški dolini. - Arheološki vestnik 42, 65-198. DULAR J., CIGLENEČKI, S. in / and A. DULAR 1995, Kučar. Železnodobno naselje in zgodnjekrščanski stavbni kompleks na Kučarju pri Podzemlj / Kučar. Eisenzeitliche Siedlung und frühchristlicher Gebäudekomplex auf dem Kučar bei Podzemelj. - Opera Instituti Archaeologici Sloveniae 1. DULAR, J., KRIŽ, B., PAVLIN, P., SVOLJŠAK, D. in / and S. TECCO HVALA 2000, Prazgodovinska višinska naselja v dolini Krke. - Arheološki vestnik 51, 119-170. EGGER, R. 1916, Frühchristliche Kirchenbauten im südlichen Noricum. - Sonderschriften des Österreichischen archäologischen Institutes in Wien 9. EGGER, R. 1950, Der Ulrichsberg. Ein heiliger Berg Kärntens. - Carinthia I, 140, 29-78. FELGENHAUER-SCHMIEDT, S. 1993, Das Kappele ("die Käpile") ob Jadersdorf. Eine spätantik-frühmittelalterliche Höhensiedlung in Oberkärnten. - Aus Forschung und Kunst 27. FIRNEIS, M. G. in / and C. KÖBERL 1989, Further studies on the astronomical orientation of Medieval churches in Austria. - V I In: A. F. Aveni (ur. I ed.), World archaeoastronomy, 430-435, Cambridge. FRASER M. 1994, Egeria and the fourth century liturgy of Jerusalem. English translation and commentary. (http:I I users.ox.ac.ukI~mikefI durhamI egetra.html [zadnji dostop I last checked 17. 10. 2008]). FURLANI, U. 1969, Una stazione militare romana sul Castellazzo di Doberdo del Lago. - Aquileia Nostra 40, 5y-y0. GABROVEC, S. 1974, Halštatske nekropole v Bohinju. - Arheološki vestnik 25 (1976), 287-318. GABROVEC, S. 1975, Ajdovski gradec. - V I In: ANSl, 164. GATTIN, N. in I and M. PEJAKOVIC 1982, Sta-rohrvatska sakralna arhitektura. - Zagreb. GESTRIN, F. 1987, Karta obsoške poti iz 17. stoletja I Eine Karte der Isonzo- Straße aus dem 17. Jahrhundert. - Zgodovinski časopis 41I1, 97-101. GINZEL, F. K., 1914, Handbuch der mathematischen und technischen Chronologie das Zeitrechnungswesen der Völker, III. Band. - Leipzig. GLASER, F. 1985, Das spätantike Gräberfeld auf dem Hemmaberg. - Carinthia I, 175, 85-89. GLASER, F. 1991, Das frühchristliche Pilgerheiligtum auf dem Hemmaberg. - Aus Forschung und Kunst 26. GLASER, F. 1993, Eine weitere Doppelkirchenanlage auf dem Hemmaberg und die Frage ihrer Interpretation. - Carinthia I, 183, 165-186. GLASER, F. 1996a, Églises doubles ou famille d'églises: les cinq églises du Hemmaberg (Mont Sainte-Hemma). - V I In: Antiquité tardive 4, 142-148. GLASER, F. 1996b, Frühchristliche Denkmäler in Kärnten. Ein Führer. - Klagenfurt. GLASER, F. 1997, Frühes Christentum im Alpenraum: Eine archäologische Entdeckungsreise. - Graz, Wien, Köln. GLASER, F. 2003, Der frühchristliche Kirchenbau in der nordöstlichen Region (KärntenIOsttirol). - V I In: R. Sennhauser 2003a, 413-437. GREEN, W. M. 1931, The Lupercalia in the fifth century. - Classical Philology 26I1, 60-69, Chicago. GUÉRANGER, D. P. 1904, The liturgical year. -London. GUNJAČA, Z. 2005, Ostaci arhitekture i analiza gradevinskih faza. - V I In: Z. Gunjača et al., Srima - Prižba. Starokrščanske dvojne cerkve, 9-54, Šibenik. GUZSIK, T. 1978, Sol aequinoctialis - zur Frage der äquinoktialen Ostung im Mittelalter. - Periodica Polytechnica: Architecture 22I3-4, 191-213. HAREJ, Z . 1989, Kozmac pri Gojačah - prazgodovinsko in poznorimsko gradišče. - Goriški letnik. Zbornik goriškega muzeja 15I16, 3-16. HAWKINS, G. S. 1965, Stonehenge decoded. -Garden City. HAWKINS, G. S. 1968, Astro-archaeology. - Vistas in Astronomy 10, 45-88. HOLWECK, F. 1908, Candlemas. - V / In: The Catholic Encyclopedia, New York. (http://www.ne-wadvent.org/cathen/03245b.html [zadnji dostop / last checked 8. 10. 2011]). HOLZ, E. 1994, Razvoj cestnega omrežja na Slovenskem ob koncu 18. in v začetku 19. stoletja. -Ljubljana. HORVAT, J. 1999, Roman Provincial Archaeology in Slovenia Following the year 1965: Settlement and Small Finds. /Rimska provincialna arheologija v Sloveniji po letu 1965: poselitvena slika in drobna materialna kultura. - Arheološki vestnik 50, 215-257. ILIC, O. 2006, Early Christian Baptistries in Northern Illyricum. - Starinar 56, 223-244. INCERTI, M. 2001, Solar geometry in Italian Cistercian architecture. - Archaeoastronomy. The Journal of Astronomy in Culture 16, 3-23. IWANISZEWSKI, S. 1989, Exploring some anthropological theoretical foundations for archaeoastronomy. - V / In : A. F. Aveni (ur. / ed.), World archaeoastronomy, 27-37, Cambridge. IWANISZEWSKI, S. 1998, The development of a regional archaeoastronomy: the case of Central-Eastern Europe. - V / In: Archeoastronomia, credenze e religi-oni nel mondo antico (Atti dei Convegni Lincei 141), 177-201, Roma. IWANISZEWSKI, S. 1999, Archaeoastronomy in traditional areas of Eurasia. A review paper. - Archaeoastronomy. The Journal of Astronomy in Culture 14/2, 87-127. KANEKAR, A. K. 1992, Celebration of place. Processional rituals and urban form. - Neobjavljena doktorska disertacija / Unpublished Doctoral Dissertation, Massachusetts Institute of Technology, Cambridge (USA). KATIC, M. 2003, Nova razmatranja o kasno-antičkom gradu na Jadranu. - Opuscula archaeologica 27, 523-528. KLASINC, R. 2001, Odlomki ločnih fibul. - V / In: P. Bitenc, T. Knific (ur. /eds.), Od Rimljanov do Slovanov. Predmeti, 67-68, št. / No. 209. KLODIČ - SABLADOSKI, A. 1912, Livško jezero. -Novo mesto. KNIFIC, T. 1979, Vranje pri Sevnici. Drobne najdbe z Ajdovskega gradca (leto 1974). - Arheološki vestnik 30, 732-385. KNIFIC, T. 1983, Bled v zgodnjem srednjem veku. -Neobjavljena doktorska disertacija / Unpublished Doctoral Dissertation, Oddelek za arheologijo, Filozofska fakulteta Univerze v Ljubljani, Ljubljana. KNIFIC, T. 1986, Ajdovski gradec pri Sevnici: poznoantična naselbina. - Arheološki pregled 27, 105-107. KNIFIC, T. 1994, Vranje near Sevnica. A Late Roman Settlement in the Light of Certain Pottery Finds. -Arheološki vestnik 45, 211-237. KNIFIC, T. 1999, Gradišče nad Bašljem. - V / In: Zakladi tisočletij, 398-401, Ljubljana. KNIFIC, T. 2004, Na stičišču treh svetov. Arheološki podatki o Goriški v zgodnjem srednjem veku. - Goriški letnik. Zbornik Goriškega muzeja 29, 5-30. KNIFIC, T. 2007, Zgodnjesrednjeveški pozlačeni predmeti z Gradišča nad Bašljem (Slovenija) / Early Mediaeval Gilded Artefacts from Gradišče nad Bašljem (Slovenia). - Prilozi Instituta za arheologiju u Zagrebu 24, 317-326. KNIFIC, T. 2010, Poznoantični Romani z Ljubinja pri Tolminu. - Tolminski zbornik 2010, 241-249. KNIFIC, T. in / and D. SVOLJŠAK 1984, Grobovi langobardskih vojščakov iz Solkana (Nove Gorice). -Arheološki vestnik 35, 277-292. KOBLER A, PFEIFER, N., OGRINC, P., TODO-ROVSKI, L., OŠTIR, K. in / and S. DIZEROSKI 2007, Repetitive interpolation: A robust algorithm for DTM generation from aerial Laser Scanner Data in forested terrain. - Remote sensing of environment 108/1, 9-23. KOKALJ, Ž., OŠTIR, K. in /and K. ZAKŠEK 2008, Uporaba laserskega skeniranja za opazovanje preteklih pokrajin - primer okolice Kobarida. - V / In: D. Perko et al. (ur. /eds.), Geografski informacijski sistemi v Sloveniji 2007-2008, 321-329. KOKALJ, Ž., ZAKŠEK , K. in / and K. OŠTIR 2011, Application of sky-view factor for the visualisation of historic landscape features in lidar-derived relief models. - Antiquity 85, 263-273. KOS, M. 1955, Zgodovina Slovencev. Od naselitve do petnajstega stoletja. - Ljubljana. KOS, P. 1986, The monetary circulation in the southeastern Alpine Region ca. 300B.C.-A.D. 1000. - Situla 24. KOS, P. 1988, Die Fundmünzen der römischen Zeit in Slowenien I, II. - Berlin. KUNAVER, J. 1998, Julijske Alpe. V / In: D. Perko, M. Orožen Adamič 1998, 54-70. KURET, N. 1972, Obredni obhodi pri Slovencih. -Traditiones 1, 93-112 KURET, N. 1989, Praznično leto Slovencev. - Ljubljana. LADSTÄTTER, S. 2000, Die Materielle Kultur der Spätantike in den Ostalpen. - Mitteilungen der Prähistorischen Kommission 35. LEBEN, F. in / and J. ŠAŠEL 1971, Hrušica - Ad Pirum. - V / In: J. Šašel, P. Petru 1971, Claustra Alpium Iuliarum I, Fontes. Katalogi in monografije 5, 93-96. LEBEN, F. in / and A. VALIČ 1978, Ajdna. - Arheološki vestnik 29, 532-545. LEBEN F. in / and Z. ŠUBIC 1990, Poznoantični kastel Vrh Brsta pri Martinj hribu na Logaški planoti. -Arheološki vestnik 41, 313-354. LOCKYEAR, N. 1894, The Dawn of Astronomy. -London. LUX, J. 2006, Poznoantična višinska naselbina Zidani gaber nad Mihovim. Geografska in arheološka predstavitev najdišč Zidani gaber in Gradec nad Mihovim. - Rast, leto 12, št. 1, 31-41. MAGGI, P. in / and B. ŽBONA TRKMAN 2007, Tra Natisone e Isonzo: il territorio in eta romana. - V / In: M. Chiaba, P. Maggi, C. Magrini (ur. / eds), Le Valli del Natisone e dell'Isonzo tra Centroeuropa e Adriatico. Atti del Convegno - San Pietro al Natisone (UD) 15-16 settembre 2006, 59-77. MAGNANI, S. 2007, Le vie di comunicazione in epoca romana / Cestne komunikacije v rimskem obdobju. - V / In: S. Magnani, A. Pesina (ur./ eds.), Terre d'incontro. Contati e scambi lungo le Valli di Natisone e dell'Isonzo dall'antichitä al medioevo. / Kraji srečanj. Stiki in izmenjave vzdolžNadiških in Soške doline od antike do srednjega veka, 129-151, Cividale del Friuli. MARCONE, A. 2004, L'Illirico e la frontiera nordorientale dell'Italia nel IV secolo d.C. - V / In: "Dall'Adriatico al Danubio. L'Illirico nell'etä greca e romana", Atti del Convegno di Cividale del Friuli, 25-27 settembre 2003, 343-359. MARTINDALE, C. C. 1909, Epiphany. - V / In: The Catholic Encyclopedia, New York. (http://www. newadvent.org/cathen/05504c.htm [zadnji dostop / last checked 8. 10. 2011]). MASELLI SCOTTI, F. 1992, Due fortificazioni tardoantiche ad oriente di Aquileia. - V / In: Felix temporis reparatio. Atti del Convegno Archaeologico Internazionale "Milano capitale dell'impero romano" Milano, 8-11 marzo 1990, 369-373. MASON, Ph. 1998, Late Roman Črnomelj and Bela krajina. - Arheološki vestnik 49, 285-313. MASON, Ph. 2006, Late Roman Bela krajina. - V / In: D. Davison, V. Gaffney, E. Marin (ur. / eds.) Dalmatia. Research in the Roman Province 1970-2001, BAR Int. Ser. 1576, 115-131. McCLUSKEY, S. C. 1993, Astronomies and ritual at the dawn of the Middle Ages. - V / In: C. L. N. Ruggles, N. J. Saunders (ur. / eds.), Astronomies and cultures, 100-123, Niwot, Colorado. McCLUSKEY, S. C. 2007, Calendrical cycles, the eighth day of the world, and the orientation of English churches. - V / In: C. Ruggles in G. Urton (ur. / eds.), Skywatching in the ancient world. New perspectives in cultural astronomy. Studies in honor of Anthony F. Aveni, 331-353, Boulder. MEIER, M. 2002, Kaiserherrschaft und "Volksfrömmigkeit" im Konstantinopel des 6. Jahrhunderts n. Chr. Die Verlegung der Hypapante durch Justinian im Jahr 542. - Zeitschrift für Alte Geschichte 51/1, 89-111. MENGHIN, W. 1985, Die Langobarden. Archäologie und Geschichte. - Stuttgart. METFORD, J. C. J. 1991, The Christian year. -New York. MICHELL, J. 1989, A little history of astro-archaeology. Stages in the transformation of a heresy. - New York. MIKULČIČ, I. 2002, Spätantike und frühbyzantinische Befestigungen in Nordmakedonien. Städte - Vici - Refugien - Kastelle. - Münchner Beiträge zur Vor- und Frühgeschichte 54. MILAVEC, T. in / and Z. MODRIJAN 2007, L'insediamento tardoantico di Tonovcov grad alla luce dei reperti ceramici e metallici / Poznoantična naselbina Tonovcov grad v luči kovinskih in keramičnih najdb. - V / In: S. Magnani, A. Pesina (ur./ eds.), Terre d'incontro. Contati e scambi lungo le Valli di Natisone e dell'Isonzo dallantichitä al medioevo. / Kraji srečanj. Stiki in izmenjave vzdolž Nadiških in Soške doline od antike do srednjega veka, 106-127, Cividale del Friuli. MILINKOVIC, M. 2001, Die Byzantinische Höhenanlage auf der Jelica in Serbien - Ein Beispiel aus dem nördlichen Illyricum des 6. Jh. - Starinar 51, 71-133. MILINKOVIC, M. 2010, Gradina na Jelici. Ra-novizantijski grad i srednjovekovno naselje. - Beograd. MILTNER, F. 1953, Die Ausgrabungen auf dem Kirchbichl von Lavant/Osttirol. - Jahreshefte des Österreichischen archäologischen Institutes in Wien 40, Beiblatt 5, 15-92. MINISTR, Z. 1997, Christian rotundas in Bohemia with pagan orientation to the Sun. - V / In: C. Jaschek in / and F. Atrio Barandela (ur. / eds.), Actas delIV. Congreso de la SEAC "Astronomia en la Cultura", 277-280, Salamanca. MIRNIK-PREZELJ, I. 1984, Ajdovski gradec nad Vranjem pri Sevnici. Arheološka izkopavanja 1980-1982. - Diplomska naloga / Diploma. Oddelek za arheologijo, Filozofska fakulteta Univerze v Ljubljani, Ljubljana. MLINAR, M. 2002, Nove zanke svetolucijske uganke: arheološke raziskave Most na Soči 2000-2001. Katalog razstave. - Tolmin. MLINAR, M. 2004, Srpenica, novoodkrito najdišče svetolucijske skupine. - V / In: Bovški zbornik, 129-142, Tolmin. MLINAR, M. in / and M. KNAVS 2004, Selce, il nuovo sito archeologico scoperto nel Tolminotto. -Quaderni Friulani di Archeologia 14, 73-79. MLINAR, M. in/ and S. PETTARIN 2007, L'eta del ferro nelle Valli dell' Isonzo e del Natisone. - V / In: M. Chiaba, P. Maggi, C. Magrini (ur. / eds.), Le Valli del Natisone e dell'Isonzo tra Centroeuropa e Adriatico. Studi e richerche sulla Gallia Cisalpina 20, 43-52. MODRIJAN, Z. 2005, Oskrba z vodo v poznoantični utrjeni naselbini Tonovcov grad pri Kobaridu / Water supply to the late antique fortified settlement of Tonovcov grad near Kobarid - V / In: A. Mihevc (ur. / ed.), Kras: voda in življenje v kamniti pokrajini / Water and life in a rocky landscape, 249-264, Ljubljana. MODRIJAN, Z. 2007, The Late Antique settlement Tonovcov grad near Kobarid: excavations 2002-2005. -V / In: M. Chiaba, P. Maggi, C. Magrini (ur. / eds), Le Valli del Natisone e dell'Isonzo tra Centroeuropa e Adriatico. Studi e richerche sulla Gallia Cisalpina 20, 175-180. MODRIJAN, Z. 2011, Continuity in Late Antiquity Slovenian fortified hilltop settlements. - V / In: O. Heinrich-Tamaska (ur. / ed.), Keszthely-Fenekpuszta im Kontext spätantiker Kontinuitätsforschung zwischen Noricum und Moesia, 157-171, Budapest, Leipzig, Ke-szthely, Rahden (Westf.). MODRIJAN, Z. in /and T. MILAVEC 2011, Pozno-antična utrjena naselbina Tonovcov grad pri Kobaridu. Najdbe / Late Antique fortified settlement Tonovcov grad near Kobarid. Finds. - Opera Instituti Archaeologici Sloveniae 24. MORRISON, L. V. 1980, On the analysis of mega-lithic lunar sightlines in Scotland. -Archaeoastronomy 2. Journal for the Historiy of Astronomy, Suppl. to Vol. 11, S65-S77. MURIALDO, G. 1992, Il castrum tardo-antico di S. Antonino nell'ambito dell'incastellmento del Finale: stato attuale delle ricerche. - V / In: G. Murialdo et al., Il "castrum" tardo antico di S. Antonino di Perti di Finale Ligure (Savona). Terze notizie preliminari sulle campagne di scavo 1982-1991. Archeologia Medievale XIX, 279-368. MUŠIČ, B. 1999, Geophysical prospecting in Slovenia: an overview from the year 1990 with some observatons related to the natural environment / Geofizikalna prospekcija v Sloveniji: pregled raziskav od leta 1990 z nekaterimi ugotovitvami glede naravnega okolja. - Arheološki vestnik 50, 349-405. NOTHDURFTER, H. 2003, Frühchristliche und frühmittelalterliche Kirchenbauten in Südtirol. - V/In: R. Sennhauser 2003a, 273-355. ODAR, B. 2006, The archer from Carnium / Lo-kostrelec iz Karnija. - Arheološki vestnik 57, 243-275. OGRIN, M. 2006, Arheološke raziskave v Julijskih Alpah. Bohinj in Blejski kot. - V / In: T. Cevc (ur. / ed.), Človek v Alpah, 96-110. ORAŽEM, F. 1989, Leto Kristusove skrivnosti. -Zbirka Božje okolje 17, Ljubljana. OSMUK, N. 1978, Nove najdbe iz časov preseljevanja narodov v spodnji Vipavski dolini. - Arheološki vestnik 29, 464-476. OSMUK, N. 1981, Bukovica. - Varstvo spomenikov 23, 232-233 OSMUK, N. 1985a, Dolenja Trebuša, Tolmin -Prvejk. - Varstvo spomenikov 27, 292. OSMUK, N. 1985b, Drežnica, Tolmin. - Gradec. -Varstvo spomenikov 27, 292. OSMUK, N. 1985c, Kobarid, Tolmin. - Stari grad (Tonovcov). - Varstvo spomenikov 27, 293. OSMUK, N. 1985d, Logje, Tolmin. - Gradec. -Varstvo spomenikov 27, 294-295. OSMUK, N. 1985e, Robič. - Molida. - Varstvo spomenikov 27, 295-296. OSMUK, N. 1985f, Sela nad Podmelcem. - Varstvo spomenikov 27, 297 OSMUK, N. 1987a, Die Bronzeplastik aus Kobarid. - Archaeologia Iugoslavica 24, 57-79. OSMUK, N. 1987b, Rimskodobna arhitektura na Ledinah v Novi Gorici. - Goriški letnik. Zbornik Goriškega muzeja 12-14, 79-104. OSMUK, N. 1997a, Kobarid od prazgodovine do antike. - V / In: Kobarid, 9-16, Kobarid. OSMUK, N. 1997b, Kekec (Sv. Katarina), Nova Gorica. - Varstvo spomenikov 36, 176. OSMUK, N. 1999, Logje. - Varstvo spomenikov 38/99, 64-66. OSMUK, N. 2001, Fibula. - V / In: P. Bitenc in / and T. Knific (ur. / eds.), Od Rimljanov do Slovanov. Predmeti, 47, št. / No. 131, Ljubljana. OVSEC, D. 1992, Velika knjiga o praznikih. Praznovanja na Slovenskem in po svetu. - Ljubljana. PAHIČ, S. 1960, Brinjeva gora. Začasen oris njene preteklosti za ljubitelje domače zgodovine. Poročilo. -Maribor. PAHIČ, S. in / and V. KOPRIVNIK 2002, Sledovi davne preteklosti. Brinjeva gora z okolico. - Muzejski listi 28, Maribor. PANCIERA, S. 1976, Strade e commerci tra Aquileia e le regioni alpine. - Antichita Altoadriatiche 9, 153-172. PAVLIN, P. in / and j. DULAR 2007, Prazgodovinska višinska naselja v Posavskem hribovju. - Arheološki vestnik 58, 65-120. PERKO, D., j. NARED, M. ČEH, D. HLADNIK, M. KREVS, T. PODOBNIKAR in / and R. ŠUMRADA (ur. / eds.) 2006, Geografski informacijski sistemi v Sloveniji 2005-2006. - GIS v Sloveniji 8. PERKO, D. in / and M. OROŽEN ADAMIČ (ur./ eds.) 1998, Slovenija. Pokrajine in ljudje. - Ljubljana. PERKO, D., M. ZORN, N. RAZPOTNIK,M. ČEH, D. HLADNIK, M. KREVS, T. PODOBNIKAR, B. REPE in / and R. ŠUMRADA (ur. / eds.) 2008, Geografski informacijski sistemi v Sloveniji 2007-2008. - GIS v Sloveniji 9. PETRIKOVITS, H. v. 1985, Duel. - V / In: Reallexikon der Germanischen Altertumskunde 6, 226-238. PETRU, P. 1971, Zarakovec. - V / In: j. Šašel, P. Petru 1971, 82-83. PETRU, P. 1972, Novejše arheološke raziskave Claustra Alpium luliarum in kasnoantičnih utrdb v Sloveniji. - Arheološki vestnik 23, 343-366. PETRU, P. 1978, Claustra Alpium luliarum und die Spätrömische Verteidigung in Slowenien. - Arheološki vestnik 29, 359-367. PETRU, P. 1979, Stavba A (episkopij?) na Ajdovskem gradcu nad Vranjem pri Sevnici. - Arheološki vestnik 30, 726-731. PETRU, P. 1980-1981, Eine bisher unbekannte Römerstrasse und die Sperrmauer auf der Hrušica (Ad Pirum). Teil des Systems Claustra Alpium luliarum. -Archaeologia lugoslavica 20-21, 132-135. PETRU, P. 1982, Arheološki oris poznoantične poselitve Slovenije. - Zgodovinski časopis 36, 295-310. PETRU, P. in / and T. ULBERT 1975, Vranje pri Sevnici: starokrščanske cerkve na Ajdovskem gradcu / Vranje bei Sevnica: frühchristliche Kirchenanlagen auf dem Ajdovski gradec). - Katalogi in monografije 12. PFLAUM, v. 2001, jermenski zaključek. - V / In: P. Bitenc, T. Knific (ur. / eds.), Od Rimljanov do Slovanov. Predmeti, 24, št. / No. 53, Ljubljana. PFLAUM, V. 2002, Spätrömische kerbschnittver-zierte Gürtelbeschläge im Gebiet des heutigen Slowenien. - Antichita Altoadriatiche 51, 259-287. PFLAUM, V. 2004, Poznorimski obrambni in vojaški sledovi 5. stoletja na ozemlju sedanje Slovenije. - Neobjavljena doktorska disertacija / Unpublished Doctoral Dissertation. Oddelek za arheologijo, Filozofska fakulteta Univerze v Ljubljani, Ljubljana. PFLAUM, V. 2007, The supposed Late Roman hoard of tools and a steelyard from Vodice near Kalce / Domnevna poznorimska zakladna najdba orodja in hitre tehtnice z Vodic pri Kalcah. - Arheološki vestnik 58, 285-332. PICCOTTINI, G. 1976, Das spätantike Gräberfeld von Teurnia St. Peter in Holz. - Archiv für Vaterländische Geschichte und Topographie 66. PIRKMAjER, D. 1994, Rifnik. Arheološko najdišče. Vodnik. - Celje. PLETERSKI, A. 2008a, Zgodnjesrednjeveška naselbina na blejski Pristavi. Najdbe / Frühmittelalterliche Siedlung Pristava in Bled. Funde. - Opera Instituti Archaeologici Sloveniae 14. PLETERSKI, A. 2008b, Kuhinjska kultura v zgodnjem srednjem veku / Küchenkultur im frühen Mittelalter. - Ljubljana. PODOBNIKAR, T. 2003, Kronologija izdelave digitalnega modela reliefa Slovenije. - Geodetski vestnik 47, 47-54. PODOBNIKAR, T. 2009, Consequences of data and information uncertainties for spatial modelling of the cultural environment. - Arheo 26. PODOBNIKAR, T., DULAR, j. in / and S. TECCO HVALA 2004, Iteractive Approach to Ancient Paths Modelling in the Iron Age Study of the Dolenjska Region (Slovenia). - V /In: Enter the Past. The E-way into the Four Dimensions of Cultural Heritage, CAA 2003, BAR International Series 1227, 254-257. RAjŠP, V. 1994, Obsoška cesta v prometni politiki avstrijske države do konca 18. stoletja. - Kronika. Časopis za slovensko krajevno zgodovino 42, 46-51. RAjŠP, V. (ur. / ed.) 1997, Slovenija na vojaškem zemljevidu 1763-1787 (1804), 3. zvezek / Josephinische Landesaufnahme 1763-1787 (1804) für das Gebiet der Republik Slowenien, 3. Band. - Ljubljana. RUGGLES, C. 1999, Astronomy in prehistoric Britain and Ireland. - New Haven, London. RUPEL, L. 1997, Rinvenimenti di superficie di materiale archeologico nel comune di Raveo. - Forum lulii 20 (1996), 35-41. RUPEL, L. 2005, Contributi alla carta archeologica dell valli del Natisone, II. - Forum lulii 29, 37-116. RUPNIK, F. 1993, Novi kobariški čudež. - Dom, leto XXVIII, št. 7, 15. 4. 1993. RUTAR, S. 1882, Zgodovina Tolminskega. - Gorica. RUTAR, S. 1890, Prazgodovinska in rimska razkopavanja po Sloveniji leta 1889. - Letopis Matice slovenske 129, 179-209. SAGADIN, M. 1987, Ajdna nad Potoki. Poznoantično naselje. - Arheološki pregled (1989), 123-124. SAGADIN, M. 1989, Ajdna. - Varstvo spomenikov 31, 248-250. SAGADIN, M. 1993, Rateče. - Varstvo spomenikov 35 (izšlo 1995), 140. SAGADIN, M. 1995, Poselitvena slika rimskega podeželja na Gorenjskem. - V I In: Kranjski zbornik, 13-22. SAGADIN, M. 1997, Ajdna nad Potoki. - Kulturni in naravni spomeniki Slovenije. Zbirka vodnikov 190. SAGADIN, M. 2000-2004, Potoki pri Žirovnici. - Varstvo spomenikov 39-41. Poročila (izšlo 2006), 155-156. SAGADIN, M. 2008, Od Karnija do Kranja. - Neobjavljena doktorska disertacija I Unpublished Doctoral Dissertation. Oddelek za arheologijo, Filozofska fakulteta Univerze v Ljubljani, Ljubljana. SANTORO BIANCHI, S. 1992, I risultati dello scavo. - V I In: Castelraimondo. Scavi 1988-199Q, I. Lo scavo. Castaloghi e monografie dei Civici musei di Udine 2, 125-228. SARIA, B. 1939, Doneski k vojaški zgodovini naših krajev v rimski dobi. - Glasnik Muzejskega društva za Slovenijo 2Q, 115-151. SENNHAUSER , R. (ur. I ed) 2003a, Frühe Kirchen im ostlichen Alpengebiet. Von der Spätantike bis in ottonische Zeit. - Bayerischen Akademie der Wissenschaften, Abhandlungen N. F., Heft 123, Bd. 1, 2, München. SENNHAUSER, R. 2003b, Frühchristliche und frühmittelalterliche kirchliche Bauten in der Diözese Chur und in den nördlich und südlich angrenzenden Landschaften. - V I In: R. Sennhauser 2003a, 9-42. SENNHAUSER, R. 2003c, Typen, Formen und Tendenzen im frühen Kirchenbauten des östlichen Alpengebietes: Versuch einer Übersicht. - V I In: R. Sennhauser 2003a, 919-980. SLABE, M. 1975, Dravlje. Grobišče iz časov preseljevanja ljudstev. - Situla 16. SLABE, M. 1982-1983, Versuch einer Darstellung der Besiedlung des Talkessels von Ljubljana. - Archaeologia Iugoslavica 22-23, 63-69. SLABE, M. 2009, Arheološke raziskave na Polho-grajski gori. - V I In: Gospod z rožo. Zbornik simpozija Ri-hard Ursini Blagaj v slovenski kulturi, 255-274, Ljubljana. SLAPŠAK, B. 1997, Starejša zgodovina Rodika. -V I In: Rodik med Brkini in Krasom: zbornik ob 35Q letnici cerkve, 19-64, Koper. SODINI, J.-P. 1984, Églises Doubles. - V I In: J.-P. Sodini, K. Kolokotsas (ur. I eds.), Aliki 2, La basilique double. Études Thasiennes 10, 255-326. STARE, V. 1980, Kranj. Nekropola iz časa preseljevanja ljudstev. - Katalogi in monografije 18. STRMČNIK-GUEIČ, M. in / and S. CIGEENEČKI 2003, Ančnikovo gradišče pri Jurišni vasi. Poznoantična in zgodnjesrednjeveška naselbina na Pohorju / Late antique and early medieval settlement on Pohorje. - Slovenska Bistrica. SUIC, M. 1976, Antički grad na istočnom Jadranu. - Zagreb. SVOEJŠAK, D. 1970-1971, Posočje v zgodnjem srednjem veku. - Arheološki vestnik 21-22, 153-162. SVOEJŠAK, D. 1975, Novejša arheološka raziskovanja na Tolminskem. - Tolminski zbornik 209-219. SVOEJŠAK, D. 1981, Most na Soči - sto let arheoloških raziskovanj. - V / In: Most na Soči 1880 -1980 - sto let arheoloških raziskovanj (razstavni katalog), 4-15. SVOEJŠAK, D. 1985, Sv. Pavel nad Vrtovinom. Rezultati sondiranj leta 1966. - Arheološki vestnik 36, 195-236. SVOEJŠAK, D. 1987, Staroselski grobovi na Eedi-nah v Novi Gorici. - Goriški letnik. Zbornik Goriškega muzeja 12-14, 105-121. SVOEJŠAK, D. 1990, Sv. Katarina nad Novo Gorico. Arheološka podoba. - Goriški letnik. Zbornik Goriškega muzeja 17, 33-55. SVOEJŠAK, D. in / and T. KNIFIC 1976, Vipavska dolina. - Situla 17. ŠAŠEE, J. 1956, Severovzhodne italske zapore v antiki. - Kronika 4, 86-93. ŠAŠEE, J. 1971, Antični viri. - V / In: J. Šašel, P. Petru 1971, 17-45. ŠAŠEE, J. 1972, Problem naseljevanja vzhodno-alpskih Slovanov. - Kronika 20, 3-6. ŠAŠEE, J. 1975, Rimske ceste v Sloveniji (viae pu-blicae). - V / In: ANSl, 74-104. ŠAŠEE, J. 1980, Zur Historischen ethnographie des mittleren Donauraums. - V / In: Die Völker an der mittleren und unteren Donau im fünften und sechsten Jh., 13-17, Wien. ŠAŠEE, J. in / and P. PETRU 1971, Claustra Alpium luliarum I, Fontes. - Katalogi in monografije 5. ŠEMROV, A. 2004, Die Fundmünzen der römischen Zeit in Slowenien V. - Mainz am Rhein. ŠKEREJ, B. 1952, Orientacija grobov v nekaterih srednjeveških nekropolah Slovenije. - Arheološki vestnik 3/1, 108-136. ŠPRAJC, I. 1991, Arheoastronomija. - Arheo: posebna številka, Ejubljana. ŠPRAJC, I. 1999, Study of astronomical alignments in archaeological sites of central Mexico: some methodological considerations. - Anthropological Notebooks 5/1, 9-29. ŠPRAJC, I. 2001, Arheoastronomija in arheologija: razmišljanja ob knjigi Clivea Rugglesa, Astronomy in prehistoric Britain and Ireland. - Arheološki vestnik 52, 385-392. ŠRIBAR, V. 1967, Zgodnjesrednjeveško grobišče in naselbina v Podmelcu - Baška grapa. - Arheološki vestnik 18, 377-388. ŠTIH, P. 1999, O vzhodni meji Italije in o razmerah ter razmerjih ob njej v zgodnjem srednjem veku. - Ge-strinov zbornik, 103-123, Ljubljana. ŠTUKL, J. 2004, Poznoantični depo s Puštala nad Trnjem / A hoard dating to late Antiquity from Puštal above Trnje. - Arheološki vestnik 55, 415-427. ŠTULAR, B. 2006a, Analiza gospodarskega zaledja v arheologiji krajin (Site catchment analysis in landscape archaeology). - V / In: D. Perko et al. (ur. / eds.) 2006, 199-210. ŠTULAR, B. 2006b, Prostor blejskih planin v srednjem veku / Raum der Bleder Almen im Mittelalter. -V/ In: T. Cevc et al., Človek v Alpah, 230-241, Ljubljana. ŠTULAR, B. 2008, Vloga kulturne dediščine pri trajnostnem razvoju varovanih območij na primeru arheološke dediščine v Kamniško-Savinjskih Alpah (The role of cultural heritage in the long-term development of protected areas: the example of archaeological heritage in the Kamnik-Savinja Alps). - V / In: D. Perko et al. 2008, 341-353. ŠUMRADA, J. 1987, O popravilu bovške ceste na prelomu iz 14. v 15. stoletje. - Zgodovinski časopis 41/2, 313-319. TAVANO, S. 1995, Notizen zur Bibliographie Mitteleuropas. - Kadmos, Der mitteleuropäische Anzeiger 8/b, 5-16. TAVANO, S. 2000, Slavko Ciglenečki, Il sito ar-cheologico di Tonovcov grad presso Kobarid. Guida, SAZU-Kobariški muzej, Ljubljana 1997, pp. 32 con ill. nel testo. - V / In: Recensioni e Segnalazioni. Memorie Storiche Forogiuliesi 80, 253. THOM, A. 1971, Megalithic lunar observatories. -Oxford. TRUHLAR, F. 1976, Problem refugijev v Sloveniji za preseljevanja. - Arheološki vestnik 27, 290-293. TSCHURTSCHENTALER, M. 2003, Lavant (Osttirol), St. Ulrich. - V / In: R. Sennhauser 2003a, 771-773. UENZE, S. 1992, Die spätantiken Befestigungen von Sadovec (Bulgarien): Ergebnisse der deutsch - bulgarisch -österreichischen Ausgrabungen 1934 -1937. - Münchner Beiträge zur Vor- und Frühgeschichte 43. ULBERT, T. 1975, Der kirchliche Baukomplex auf dem Hügel von Vranje. - V / In: P. Petru, Th. Ulbert, Vranje pri Sevnici. Starokrščanske cerkve na Ajdovskem gradcu. Katalogi in monografije 12, 21-77. ULBERT, T. 1979a, Vranje bei Sevnica. Siedlungsgeschichtliche Untersuchungen. - Arheološki vestnik 30, 695-725. ULBERT, T. 1979b, Zur Siedlungskontinuität im südöstlichen Alpenraum (vom 2. bis 6. Jahrhundert n. Chr.) Dargestellt am Beispiel von Vranje (ehem. Untersteiermark). - Vorträge und Forschungen 25, 141-157. ULBERT, T. 1981, Das spätrömische Kastell AD PIRVM - Hrušica. - V / In: Ad Pirum (Hrušica). -Münchner Beiträge zur Vor- und Frühgeschichte 31. ULBERT, T. 1988, Zur liturgisch-funktionellen Ausstattung spätantiker Kirchen des Alpenraums. - V / In: Die Bajuwaren. Von Severin bis Tassilo 488 - 788. Gemeinsame Landesausstellung des Freistaates Bayern und des Landes Salzburg, 287-292. UŠENIČNIK, F. 1945, Katoliška liturgika. - Ljubljana. VALIČ, A. 1968, Gradišče nad Pivko pri Naklem. -Arheološki vestnik 19, 485-508. VALIČ, A. 1981, Ajdna nad Potoki. - Varstvo spomenikov 23, 266-269. VALIČ, A. 1985, Ajdovska luknja pri Soteski v Bohinju, Radovljica. - Varstvo spomenikov 27, 272-274. VALIČ, A. 1997, Arheološki drobci vojaških prvin pozne rimske dobe na Gorenjskem. - Arheološki vestnik 48, 261-268. VIČIČ, B. 1983, Die Geräte von Ajdovski Gradec oberhalb Vranje bei Sevnica. - Balcanoslavica 10, 71-77. VIDRIH PERKO, V. in / and M. ŽUPANČIČ 2003, Il popolamento della Slovenia sudoccidentale e dell'Istria settentrionale nel periodo tarodomano e nell'alto medio evo alla luce delle ceramiche di importazione. - V / In: Histria Antiqua 11, 457-476. VIDRIH PERKO, V. in/ and M. SAGADIN 2004, Gorenjska v antiki. - Kamniški zbornik 17, 207-224, Kamnik. VILLA, L. 2001, Nuovi dati archeologici sui centri fortificati tardoantichi - altomedievali del Friuli. - V / In: Paolo Diacono e il Friuli altomedievale (secc. VI-X). Cen-tro italiano di studi sull'alto medioevo Spoleto, 825-861. VILLA, L. 2006, Un insediamento fortificato goto sul colle di San Giorgio ad Attimis. - Forum Iulii 30, 219-221. VITRI, S., VILLA, L. in / and A. BORZACCONI 2006, Trasformazioni urbane a Cividale dal tardoanti-co al medioevo: spunti di riflessione. - Hortus artium medievalium 12, 101-122. VINSKI, Z. 1980, Ovrednotenje grobnih pridatkov. - V / In: V. Stare, Kranj. Nekropola iz časa preseljevanja ljudstev. Katalogi in monografije 18, 17-32. VOLARIČ, T. 2008, Pustovanje: ločnice in prehodi. - Traditiones 37 (1), 47-78. VUGA, D. 1970, Kozlov rob nad Tolminom. - Varstvo spomenikov 15, 172. VUGA, D. 1979, Podmelec. - Varstvo spomenikov 22, 291. WEBER, F. P. 2004, Zu den Problemen der kirchlichen Terminologie in einigen romanischen Sprachen. -Zeitschrift für Romanische Philologie 120, 231-265. WERNER, J. 1962, Die Langobarden in Pannonien. Beiträge zur Kenntnis der langobardischen Bodenfunde vor 568. - Abhandlungen der Bayerischen Akademie der Wissenschaften, philosophisch - historische Klasse, Neue Folge 55. WERNER, J. 1992, Golemanovo Kale und Sadov-sko Kale. Kritische Zusammenfassung der Grabungsergebnisse. - V / In: S. Uenze 1992, 391-417. WISEMAN, T. P. 1995, The god of the Lupercal. -The Journal of Roman Studies 85, 1. WOLFF, H. 2000, Die Frage der Besiedlung des heutigen Slowenien im Lichte des Anonymus von Ravenna. - V / In: R. Bratož (ur. / ed), Slovenija in sosednje dežele med antiko in karolinško dobo: začetki slovenske etnogeneze / Slowenien ind die Nachbarländer zwischen Antike und karolingische epoche: Anfänge der slowenischen Ethnogenese. Situla 39, (Razprave 18), 97-106. ZAVRTANIK, J. 1982, Nova Gorica. - Varstvo spomenikov 24, 187. ZAVRTANIK, J., 1984, Poznoantično grobišče pri Kosovelih. - Goriški letnik. Zbornik Goriškega muzeja 11, 85-94. ZUPANČIČ, M. in /and T. WRABER 1989, Fito-cenologija. - V / In: Enciklopedija Slovenije, 3. zvezek, 118-120, Ljubljana. ŽBONA-TRKMAN, B. 1981, Kanal. - Varstvo spomenikov 23, 272. ŽBONA-TRKMAN, B. 1987, Memento. Ajdov-ščina-Castra-Fluvio Frigido. - Goriški letnik. Zbornik Goriškega muzeja 12-14 (1985-1987), 315-319. ŽUPANČIČ, M. 1991, Poznorimsko najdišče Molida pri Robiču. - Goriški letnik. Zbornik Goriškega muzeja 18, 166-169. Založba ZRC/ ZRC Publishing http://zalozba .ac-sazu. si ISSN 1408-5208