BOUGHT WITH THE INCOME FROM THE SAGE ENDOWNENT FUND THE GIFT OF ilcnrg W. Sage 1891 A-MTim S/m Cornell University Library DF 261.A2W8S Aetolia; 3 1924 028 301 962 *...i Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924028301962 AETOLIA HENRY FROWDE, M.A. PUBLISHER TO THE UNIVERSITY OF OXFORD LONDON, EDINBURGH, AND NEW YORK AETOLIA ITS GEOGRAPHY, TOPOGRAPHY, AND ANTiaUITIES WILLIAM J. WOODHOUSE, M.A., F.R.G.S. CLASSICAL LECTURER IN THE UNIVERSITY COLLEGE OF NORTH WALES, BANGOR SOMETIME CRAVEN FELLOW IN THE UNIVERSITY OF OXFORD FORMERLY STUDENT OF THE BRITISH SCHOOL AT ATHENS WITH MAPS AND ILLUSTRATIONS' AT THE CLARENDON PRESS 1897 Ir PRINTED AT THE CLARENDON PRESS BY HORACE HART, M.A. PRINTER TO THE UNIVERSITY PREFACE With the exception of Chapters VI, XXI, XXII, and the whole of the Notes, this book is in the main identical with the Essay to which, in 1894, was awarded the Conington ERRATUM Page 127, 1. 9 from top, /o/" Old Aetolia read Old Pleuron poetical reminiscence are beginning to fail by the time that Aetolia is reached. Very striking is the contrast between the splendid remains with which Akarnania is studded, and the miserable fragments of the poorer art on the east of the Acheloos. The simplicity of the topography of Aetolia, the mutilation of its ruins, and their technical characteristics, are explicable by the light of the history of the country : at the same time that history only attains its full value by borrowing actuality from geography and topography. Should any one be tempted to look askance upon the latter subject, let him recall the words of Professor Ramsay ^ : — ' I quote from his Historical Geography of Asia Minor (p. 51 fol.) — a thesaurus of knowledge, and a very text-book of topographical method. PREFACE With the exception of Chapters VI, XXI, XXII, and the whole of the Notes, this book is in the main identical with the Essay to which, in 1894, was awarded the Conington Memorial Prize offered in Oxford 'once in every, three years for a dissertation ... on some subject appertaining to classical learning.' It represents the chief part of my work in Greece during the years 1892 and 1893 as Craven Fellow. It is at first sight somewhat surprising to find that the topography of Aetolia is so little complicated. It presents few great problems. Nor, as a rule, are the ruins them- selves attractive. There is little in them to interest the non-scientific observer, and ordinary stores of historical or poetical reminiscence are beginning to fail by the time that Aetolia is reached. Very striking is the contrast between the splendid remains with which Akarnania is studded, and the miserable fragments of the poorer art on the east of the Acheloos. The simplicity of the topography of Aetolia, the mutilation of its ruins, and their technical characteristics, are explicable by the light of the history of the country : at the same time that history only attains its full value by borrowing actuality from geography and topography. Should any one be tempted to look askance upon the latter subject, let hiin recall the words of Professor Ramsay^ : — ' I quote from his Historical Geography of Asia Minor (p. 51 fol.) — a thesaurus of knowledge, and a very text-book of topographical method. Viii PREFACE ' Topography is the foundation of history. No one who has familiarized himself with Attic history in books, and has afterwards ascended Pente- licus and seen that history spread forth before him in the valleys and mountains and sea that have moulded it, will ever disbelieve in the value of topography as an aid to history. . . . Yet few that study Greek history, and play the part of examiner or examinee in it, realize what we owe to the greatest of modern topographers, Leake. Who, that goes through the usual course of highest honours in ancient history and literature, ever hears the fame of Leake, or knows that he has done more to make a real understanding of Greek life possible than any other Englishman of this century ? We all know Curtius' History of Greece : how many of us know a finer and greater work, Curtius' Peloponmsos ? Some of us are even so narrow as to imagine that the reading of some modern books, supplemented by a little study of Thucydides, Herodotus, and Xenophon (a few reach Polybius— how very few go deeper !), will enable us to understand ancient history. If we want to understand the ancients, and especially the Greeks, we must breathe the same air that they did, and saturate ourselves with the same scenery and the same nature that wrought upon them. For this end correct topography is a necessary, though a humble, servant.' Our predecessors in the field of Aetolian topography practically reduce themselves to two, — Colonel Leake and M. Bazin. For the older writers belong to a different order. Compelled to forgo, or even scarcely dreaming of, living contact with the land itself, they are unable to distinguish the wheat from the chaff in their gleanings from the Classics. At an early stage of their investigations they find themselves involved in hopeless confusions and contradictions, which baffle every attempt at re-construction. And Pouqueville's claim to be taken seriously as a topographer is disposed of by his own nai've confession, — 'je donnai par une sorte d'inspiration des noms a tous les lieux qui m'environ- naient.' How must we regard one who, with respect to the facts of his subject, exhibits the recklessness found so often in the pages of the Voyage dans la Grecel We despair ol dissecting out genuine observations from the tissue of fiction and conjecture\ A new era begins with the publication of Leake's Travelc ^ Cf. Puillon-Boblaye, Expedition Scientijique de Marie, p. lo: 'Pouquc- ville qui malheureusement, suivant Mannert, repand tant de fleurs sur sa route qu'il est quelquefois difficile d'en reconnaltre la trace.' PREFACE IX tn Northern Greece. In dealing with Aetolia the 'model traveller' is certainly less happy than in the case of any other section of Hellas. Doubtless, if he could have ex- plored the country more thoroughly, both his own errors and the discoveries reserved for his successors would have been reduced in number. Leake's acquaintance with Aetolia was indeed extrenaely limited. From Patras he sailed to Varassova and proceeded to Mesolonghi. Thence he crossed the Zygos to Vrachori, visiting the kastro of Kyria Eirini on the way. From Vrachori he made an excursion to the ruins at Paravola (Kuvelos) and Vlochos, and crossed the Acheloos into Akarnania. On his second visit he reached Aetolia from the west, and traversed the Paracheloitis : on his way from Mesolonghi to Bochori he inspected the ruins of Kaly- don. In Eastern Aetolia he made but a "single journey, from Amphissa by way of Lidhoriki and the range of Trikorpho and Vigla to Naupaktos. With regard to the number of sites discoverable upon Leake's routes we have hardly anything to add to what we find in the Travels. We are principally engaged in correct- ing the slight inaccuracies and in filling up the gaps that inevitably mar the notes of a traveller when he examines a site for the first time, often under the pressure of the hundred and one necessities of the road. With respect to the identifications also Leake left little for subsequent explorers to do. We marvel at the precision with which he assigned their names to the remains: Chalkis, Kalydon, Halikyrna, New and Old Pleuron, Pylene, Proschion, Paia- nion, Konope, Lysimacheia, Trich onion,— all these were identified by Leake. With far other and deeper meaning might he have exclaimed with Pouqueville that by inspira- tion he named the ruins in his path. It was an inspiration born of an intimate knowledge of the ancient texts and a genuine feeling for the conditions of fife in classical times, both combined with a soldier's eye for geography. Of a different stamp is Bazin. He reminds us forcibly of Pouqueville. We find in the Memoire sur I'Etolie that slightness of treatment which we have remarked as charac- X PREFACE teristic of the Voyage dans la Grece. Its inequality also made it impossible for us to regard the Memoire as final. To a certain extent, perhaps, Bazin is not to blame for this defect, as travelling in Aetolia must have been more difficult thirty years ago than it is now. Still, the fact remains, that only in Old Aetolia, of which Leake had all but exhausted the possibilities fifty years before him, is Bazin really com- plete. It is strange to find so many sites unknown to, him, even in Central Aetolia where his work was most original. In the matter of identification Bazin, of course, takes us a long way further than was possible for Leake. To his credit fall the identifications of Phana, Elaos, Phytaion, Boukation, Krokyleion, and Teichion. A special merit of his Memoire is the perception of the importance of the technical varieties exhibited in the masonry of the ruins, and the endeavour to define them precisely. In the following pages, while taking the strictly topo- graphical part a stage beyond the point reached by previous inquirers, an attempt has been made to produce something more than a mere list and description of the sites with conjectures as to their ancient names. However imperfectly, I have tried to direct attention to the physical conditions and the natural relations under which the towns of Aetolia stood, and to trace the influence of these factors upon the part played in history by their inhabitants. Furthermore, I can feel no sympathy with the conservatism that is content to pore for ever over a printed text while contemning the fragments of reality that almost daily are revivifying the dead page and filling up its lacunae. I have endeavoured, therefore, to supplement the meagreness of the literary sources of information by an appeal to every particle of evidence discoverable on the sites themselves. To compare small things with great, I have tried to do for Aetolia what Professor Ramsay has done for the cities of Phrygia, — how rich the stores of literary and epigraphic material in the one case, in the other how poverty-stricken! Yet the Aetolian cities are bound together in the pathos of a truly national history. That to some this supplementary evidence should PREFACE XI appear to have been unduly strained, and the conclusions to which it leads purely fanciful, is inevitable : but it is only by expressly formulating such conclusions that antagonism can be aroused, energy be directed upon a definite point, and a net result of truth be carried away to the sum of know- ledge. In the above statement of aims there is implied the desire to make this treatise the second member of a triad. It presupposes what we may call a 'History of the Art of Fortification in Greece,' a work unfolding the principles governing military construction in Aetolia and Greece gener- ally, and justifying the conclusions drawn therefrom, or at least the very attempt to draw conclusions. Secondly, since it is only at the bidding of the Muse of History that the dry bones of topography become instinct with . life, we must proceed from the description of the Aetolian land to the story of the Aetolian people. The history of Greek mili- tary engineering has never yet been seriously attempted; and the history of the Aetolian League requires to be re-written. 'These things lie upon the knees of the gods': their accomplishment is perhaps reserved for others, better equipped and qualified for the task. It is at any rate high time to have done with those hap-hazard and thoroughly unscientific generahties which pass muster as descriptions of Hellenic ruins. What is required in order to satisfy the demand for actuality and contact with life in classical studies is that the Hellenic kastra should all be minutely and lovingly investigated and drawn on a large scale by men thoroughly equipped for the work and trained to inter- pret tTie historical story told by the mute eloquence of the stones. This task, if undertaken at all, must be undertaken without delay, as each day sees the process of destruction more advanced. My best thanks are due to the Delegates of the Uni- versity Press for accepting my work. To two of their number a more special debt of acknowledgement must be paid. The Vice-Chancellor, the Rev. J. R. Magrath, D.D., xii PREFACE Provost of Queen's College, has been indefatigable in watch- ing over the practical details of the publication : it is largely the result of his warm patronage that the book sees the light in its present form. Mr. D. B. Monro, Provost of Oriel, has shown unwearied kindness in revising my MS. Scarcely a page but bears the traces of his ripe literary experience, — 'cerulas enim tuas miniatas illas extimescebam.' I must thank the Council of the Society for the Promotion of Hellenic Studies for a generous Grant towards defraying the cost of Illustrations, and the Oxford Craven Committee for similar aid. Mr. and Mrs. Gordon-Oswald of Aigas have also a great claim upon my gratitude. Nor may I omit the name of the Rev. H. F. Tozer, whose interest in my work as Craven Fellow has fanned an enthusiasm kindled by his own writings. My many Greek friends also claim the tribute of remembrance here. The name of almost every village in Aetolia recalls hospitality and kindly service rendered by Demarch, priest, agoghiat, shepherd, and soldier, many of whom now lie on the slopes of Mount Othrys, — TeSva/ievai yap Ka\6v evl ■npoji.a)(Oun Treirovra avhp ayadoVf wepl ^ Trarpidi p^pvdpevov. Finally, I must acknowledge my indebtedness to the skilful labours of the staff of the Clarendon Press. W. J. W. Sedbergh, Yorkshire : 1897. CONTENTS 3 GEOGRAPHY. CHAP. I. Introductory II. The Aetolian Plains 9 III. The Great Watershed 20 IV. Agrapha . 25 V. Kravari 40 VI. Conclusion 44 VII. The Aetolian Tribes, I. Old Aetolia, Apodotoi, Ophioneis, Kal- LiEis, Bomieis 55 VIII. The Aetolian Tribes, II. Aperantoi, Agraioi, Eurytanes, Thestieis 72 TOPOGRAPHY AND ANTIQUITIES. HOMERIC AETOLIA. IX. The Coast Plain, I. Kalydon, Chalkis, Halikyrna , . 91 X, The Coast Plain, II. Old and New Pleuron, Olenos , . 115 XI. Sites on the Zygos. Pylene, Proschion, Phana, Elaos . . 137 XII. The Paracheloitis. Ithoria, The Forts, Paianion, The Lakes 150 central aetolia. XIII. Sites between the Aspro and the Eremitsas. Agrinion, Mavrovru, Temple on the Zer- VAS, Palai6pyrgos, The Eiteaioi and Eoitanes 169 XIV CONTENTS CHAP. XIV. XV. XVI. XVII. XVIII. Sites on the North of Lake Trichonis. Vloch6s, Parav6la, SoBONfKos, Kryo- NER1&, ProstovAs, H. TaxiArchis, BerI- Kos, Guritsa, M6kista .... 182 Sites on the South of the Lakes : the Western Section. Konope-Arsinoe, Lysimacheia, The two Lakes 209 Sites of the Apokuro, I. Trichonion, Phytaion, Makryn6, Kal6dhi, Mesovi^ni, DHERviKisTA, Mor6sklavon, Petroch6ri, Kephal6vryson, Chryso- vfTSA 228 Sites of the Apokuro, II. Metapa, Pamphia, Thermon, Akrai, Ello- PION 249 Thermon. A Chapter of Criticism .... 263 NORTHERN AETOLIA. XIX. North Aetolia. Kol6pyrgos, Aperanteia, The Kampylos, Oracle of Odysseus, Oichalia eastern aetolia. XX. Naupaktos, and the Sites in its vicinity. Naupaktos, Molykreion, Makynia, The Bouttioi XXI. Sites on the East of the Mornos. OiNEON, EUPALION, ErYTHRAI, The Ne- MEioN, Antikyra, Potidania, Apollonia, Krokyleion, Teichion .... XXII. Ophioneia. Aigition, Kallion, Mount Korax APPENDICES. I. Towns and Mountains not identified II. Modern Marching III. The SkAla Inscriptions IV. Authorities consulted in the work . INDEX 287 309 339 363 377 379 380 385 389 XV MAPS AND ILLUSTRATIONS General Map of Aetolia To face page t. Geological Map of Aetolia 45 Map of Ancient Aetolia 53 The Lagoons of Lower Aetolia 167 Central Aetolia 239 Sketch Map of Central Aetolia (to illustrate Colonel Leake's theory of Philip's Route in 218 B.C.) 265 Sketch Map of Central Aetolia (to illustrate Becker's theory of Philip's Route in 218 b.c.) . ' 273 Sketch Map of Central Aetolia (to illustrate M. Bazin's theory of Philip's Route in 218 b. c.) 277 The Kleisiira ; looking North 14 Agrinion (Vrachori) ; looking South to the Zygos .... 17 Gorge of the Acheloos ; looking up the river from the Bridge of Tatarna 25 The Pass of Prossos ; looking North up the Karpenisi River . . 28 Karpenisi, with the Peak of Veliichi in the distance ; from the West 30 Monastery of Tat^ma ; from the South- West 35 The Bridge of Tatdrna ; from the East 36 End of the Kleisiira : Lagoon of Aetolik6 in the distance ... 49 Pteri Mountains and River of Agrapha, with the Bridge of Man6lis 81 Northern wall of Kalydon ; looking West 99 Wall of Chalkis, showing mode of construction ; from the interior of the Fortress no Tower of Chalkis, and View over the Corinthian Gulf: The Kakf- skala to the left 112 Tower and Wall of New Pleuron ; from the South-East . . • 115 Fig. I. New Pleuron. Second Tower north of Main Gate, with steps leading into Tower 116 Gate in Eastern Wall, New Pleuron ; looking East . . . • 117 Fig. 2. New Pleuron. Plan of Main Gateway 118 Fig. 3. New Pleuron. Plan of Theatre 119 Figs. 4, 5. New Pleuron. Great Cistern. Plan and Elevation . 120 The Great Cistern of New Pleuron ; from the South . . .121 Agora of New Pleuron ; from the South 122 Hill and Village of Hdghios Elias at the Almonds ; looking South . 154 xvi MAPS AND ILLUSTRATIONS Fort of St. John near the Acheloos ; looking South . To face page i6o Part of Eastern Wall of Ancient Agrinion 17° Fig. 6. Temple on Z6rvas. a. Plan. b. Section of Side-wall . 175 Palaiopyrgos ; from the North ^7° Fig. 7. Inscription on Boundary Stone, Eremitsas . . • • ^^^ Great Gate of Vlochos ; from without. Part of Mount Viena in the distance -"-^ Fig. 8. Vlochos. South-West Gateway 188 Fig. 9. Paravola. West Gateway ^9^ Fig. 10. Paravola. Semi-circular Tower 192 Semi-circular Tower at Paravola ; from the North .... i93 Interior Face of Town Wall at Paravola ; from the South . . 194 Fig. II. Inscription from Kryoneru i99 Churches of Mokista ; from the South- West 206 Ruined Church of St. George, Anghelokastron ; looking East . . 211 Fig. 12. Inscription from Mokista 213 Figs. 13, 14. Inscriptions on clay slabs, Anghelokastron . . . ibid. Fig. 15. Inscriptions from Gavalii 234 Terrace at Chrysovitsa ; looking North-East 247 Lake Trichonis, with the hill of Vloch6s and Mount Viena in the distance ; from the neighbourhood of Matardnga . . . 266 Site of Phana ; from the North 281 Bronze Thumb from the Palaio-bazari, Kephalovryson (Nat. size) . ibid. Fig. 16. Inscription from the Palaio-bazari, Kephalovryson . . 285 Kolopyrgos ; from the East 289 Fig. 17. Slab in threshing-floor, Haghios Vasileios .... 291 Fig. 18. Sketch Plan of Gateway, Djuka 293 Fig. 19. Profile of Stylobate, Temple at Velvlna .... ibid. Gateway of Djuka ; looking South-West 294 Wall at Vtilpi ; from the South 296 Port of Naupaktos; from the Second Cross -wall of the Kastro . 310 Fig. 20. Inscription from the Asklepieion of Naupaktos . . . 313 Fig. 21. Velvlna. Plan of Temple and Stoa 325 Kastro of Mamdku. Wall and Tower at North- West corner of the lines 326 Temple at Velvlna ; from the South-East 329 Fig. 22. Inscription froni the Asklepieion, LongS .... 333 Fig. 23. Inscribed slab at Sules 347 Plain of the Mornos: Antirrhion and Mount Kl6kovain the distance 353 Kastro of Veliichovos. Part of Southern Wall 363 Fig. 24. " Inscribed Block near the Sten6, Veliichovos . . . 365 Fig. 25. Inscribed Block near the Sten6, Veliichovos . . . 375 Fig. 26. Inscription from a Stele near the Phldharis . . . ibid. Fig. 27. Inscribed Stele from the neighbourhood of Lobotina . ibid. Missing Page GEOGRAPHY "Iva Se lir) T&v rdiTiav ayvoov- fievav aWTToraKTa Koi Kto^ta ■yiVijrai to Xey6iJjeva, avvinro- heiKriov hv eu] Trjv cjyia-iv Koi Ta^iv avT&v, 6 Sri Koi wap o\r]v Tqv TT pay flare Lav Tretp&fieoa iroieiv. Pol. V. 2T. CHAPTER I. Introductory. There is a striking contrast between the two shores of the Corinthian Gulf. The characteristic feature of Achaia^ is the belt of fertile land that falls from the mountains to the sea in a double terrace of varying breadth. The, outermost terrace is a narrow irregular fringe, deposited by the numerous torrents which rush down in short courses at right angles to the coast-Une and carve into cliffs and plat- forms the intervening hills of marl. Behind the marl, which constitutes the second tier, rise mountains of conglomerate. These gradually approach the Gulf as we advance westwards, and the marl formation is proportionately reduced in width, until at last the conglomerate rudely bursts through it to the sea. How different the northern shore ! From the white summits of Parriassos in the east, we behold nothing but mountains stretching westwards until they end opposite Patras in the two bold rocks of Klokova and Varassova. Far into the interior of Aetolia the picture is the same, — that of a vast complex of mountains, in which here and there a glistening peak stands out carrying its burden of snow long after winter has fled from the lowlands. And on this coast the hills, almost entirely composed of limestone, rise straight from the water. Denudation has long ago deprived them of their earthy covering, which, however, has been too scanty to form a continuous strip of flat land along the sea. The immediate consequences are obvious. From Corinth to Patras we journey as through a garden, passing in quick succession village ,and town surrounded by the most luxuriant vegetation. Every species of fruit-tree known to ' See Philippson, Der Peloponnes, i. 136-155 ; especially p. 150. B 3 4 INTRODUCTORY. [cH. Greece flourishes there in profusion, and the graceful poplar and the dark beauty of the cypress give to the landscape that charm which is sadly waiiting to so much Greek scenery. On the opposite coast we are confronted by an arid waste of mountain-land, apparently almost without inhabitant. Here and there, as we round a rocky headland, we see a village on the slopes. Only where the Mornos has deposited the spoils brought from the interior do we find any considerable space between the mountains and the sea ; and there is Naupaktos, the only large town between Galaxidhi and Mesolonghi. As we contemplate the contrasting pictures what a flood of fight is shed upon the pages of history ! There, on the right hand and the left, are outspread in mountain and plain the causes that determined the broad lines of historical develop- ment on the two shores of the Gulf. Perhaps in none better than in the history of ancient Greece is the influence of geographical conditions realized. There had not yet arisen that extreme complexity in national and civic relations which tends partly to obscure, and partly to substitute itself for and to oppose, the external influences summed up in the word 'geography.' History is there seen reduced almost to its simplest elements, and we can estimate with some accuracy the value of each factor of the product. With no wavering finger has Nature traced the boundaries of Aetolia. On the west the broad white bed of the Aspro- potamos, the Acheloos of ancient Greece, cleaves a line of demarcation from north to south, from Epiros to the Ionian sea^ On the east the range of Pindos runs down from Macedonia along the frontier of Thessaly from north-west to south-east. Typhrestos, towering at the head of the Sper- cheios valley, stands sentinel on the threshold of North Aetofia. The Oxya hills ^ prolong the line towards the ' By some the Corinthian Gulf was considered to begin from the mouth of the Acheloos. Cf. Strabo, p. 335 : 'o 8e 'S.opwBiaK.os koXttos SpxeToi fiiv arro tS>v sk^oXSiv tov Ev^vov, TtKer Be i^aaiv toC 'A;(eXc3ov, (cal tov 'Apa^ov. See also id. p. 336 ; and p. 450 : TA AtrtaXiKov xXuffTai rij n-oiovcri} flaXarrg top Kopiv6ia<6v koXttov, els ov Koi 6 'A^eX^or nora/ios e^ii/crtv. ^ 'O^d, 'Ofuai'r ; from the beech-trees (o|uai) which grow upon it. Leake (N. G. ii. 18) calls the range Oxies. I] INTRODUCTORY. 5 south-east, to Vardhusi and Ki6na\ loftiest of all the mountains of Greece. Then, south of the river Mornos, Trikorpho and Vigla^, sinking slowly down to the Gulf, complete the long chain that shuts off Aetolia from her eastern neighbours. The space enclosed within these frontiers is roughly triangular : the coast line forms the base, and the apex falls in the northern part of the canton of Agrapha. The struc- ture of the area proves on examination to be much more simple and regular than we should expect. North and South Aetolia are in fact distinguished from each other in a well-marked and unmistakable way. The natural lines in South Aetolia run east and west, while those of North Aetolia run north and south. The dividing line falls a little more than half-way between the base and apex of the triangle ; that is to say, between the coast-line and Agrapha. It coincides, therefore, with the course of the Agalianos river, which rises in the Oxya hills and flows westwards to the Acheloos. Looking first at the southern division, we notice at once the river Phidharis which in a rough way bisects it dia- gonally. The course of the river falls into three strongly marked and nearly equal sections. Rising almost upon the eastern verge of Aetolia, it flows first towards the west, gradually trending away to the south-west, until it turns suddenly due south. In this direction, cleaving the very centre of Aetolia, it flows for again nearly a third of its course. A second turning-point, almost as sharp as the first, marks the resumption of its former direction, i. e. towards the south-west, until it falls into the sea at a point roughly mid- way between the mouths of the Aspro and the Mornos. The Mornos, the third large stream of Southern Aetolia ^ resembles the Phidharis in its zig-zag course. It rises ' Few Proper names are written in so many ways in books of Greek travel as Kiona. The French map calls the mountain Guiona. Ulrichs writes the name Jona. The variety seems to arise from the attempt to reproduce the popular pronunciation, TKiava. Those are wrong who write- Khiona or Chiona, as though the word were connected with j^iiiv, ' snow ' ; See Becker, Diss. iii. p. 5, and note 14. '' TpUufxjjos (fr. Kopv(j)Ti). Cf. Tpeis Ke0a\ai in Mount Kithairon : Corfu, (='s Tovs (copious). BiyXa = Watch-tovver. ^ The other two are, of course, the Aspro and the Phidharis. 6 INTRODUCTORY. [ch. between Vardhusi and Kiona, and flows due south for about one third of its total length ; then with a sharp bend it runs to the south-west, only to turn a second time and to resume its former line of direction, thus finally reaching the Corin- thian Gulf on the east of Naupaktos. The upper Mornos \ therefore, is parallel to the central section of the Phidharis ; and the upper Phidharis to the central section of the Mornos. Thus the two rivers enclose a quadrangular space, of which the boundaries on north aiid west are formed by the Phidharis, and those on east and south by the Mornos. Within this space the eye searches in yain for a level spot : it is the rude district of Kravari. Let us now look at the coast-line. It would form a^long irregular level strip, extending from the mouth of the Aspro to that of the Mornos, were it not for those two striking precipices which tower from the sea over against Patras. On the east Mount Klokova has a height of 1041 metres ; on the west Mount Vardssova rises 917 metres sheer from the Gulf. These two rocks, separated from each other by the vale of Gavrolimni, form two short ridges running from north to south between the Phidharis and the sea, isolated masses of limestone that sever the Venetiko from the district of Mesolonghi. A rough hill country of sandstone formation ex|ends from them north-eastwards to the great pyramid of Mount Rhigani which overshadows Naupaktos. Towards the sea the interval between Klokova and Rhigani is occupied by a narrow strip of level fertile land which runs out in the long point of Antirrhion opposite the similar cape of Rhion on the Achaian side of the Gulf. Disregarding the projection of Rhigani and the twin rocks of Vara,ssova and Klokova, the first zone of Aetolia may be described as a narrow plain, never more than four, miles wide, and generally very much less, almost wholly formed by fluviatile deposits; and, therefore, widening towards the west, in correspondence with the superiority in size of the Aspro over the Phidharis, and of the latter river over the Mornos. .East of the Mornos estuary the fertile strip ceases altogether, and the mountains come quite down to the sea until we reach the gulf of Silona, where the Pleistos and the streams of Amphissa have created the rich ^ Or that part of it which is called the M^ga. See p. 42. I] INTRODUCTORY. 7 plain that belonged in ancient times to Kirrha and Krissa. The belt of flat land along the Aetolian sea-board lies nearly east and west ; but from Mesolonghi it naturally runs north- wards along the Acheloos until it is stopped by the spurs of the Zygos abutting upon the river. A second zone is formed by the Zygos range itself, extending from the Aspro to the lower angle of the Phid- haris. It does not stop there : the line sweeps upwards to Vardhiisi in the north-east through the heart of Kravari, forming the mountains known as Makr^voros, Papadhia, Tritzovon, and Vlachoviino, — in that order, enumerating them from west to east. The Mornos flows along their southern foot, in the narrow valley between them and the mountains Trikorpho and Vigla on the left bank of the stream. Crossing the Zygos we descend into a third zone, the central plain of Aetolia, in which lie the two lakes. This plain stretches from beyond the Acheloos alniost as far as the Phidharis, but a series of hills separates it from the latter river. On the left bank of the Phidharis, however, the configuration is the same, the vale of the Kakavos corre- sponding to the central basin. North of the Kakavos we have again a mountain zone, the range of Platanos, — Mounts Ardhini and Trekuri. Along the northern foot of these hills flows the upper Phidharis. Finally, from the Acheloos right across to the Oxya range on the eastern confines of Aetolia there runs the great water- shed of Kiitupas and Arabokephalon. Southern Aetolia, therefore, falls into a system of parallel lines, alternate mountain-ridge and valley, bisected by the Phidharis. On the western side of the river the direction of the lines is east and west : on its eastern side they have an upward tendency towards the north-east. Another point to be noticed is that, in addition to being roughly parallel, the ridges rise in tiers, their altitude gradually increasing as we move northwards from the coast. Klokova is 1041 metres above the sea level; Rhigani 1475 metres, Trikorpho 1552 metres. In the second zone we have Papadhia, 1714 metres, and Tritzovon, 1736 metres. In the third, Trekuri rises to 1787 metres. The loftiest summits of Arabokephalon tower to a height of 1823 metres and 1927 metres. Behind Arabokephalon, Chelidhoni reaches an altitude of 1980 8 INTRODUCTORY. metres, and Kaliakiidha of 2104 metres. Veliichi, the most northerly, is also the highest point, 2318 mfetres above the sea. It remains for us to clothe with more detail this skeleton of the territory under examination. It is clear that we have already in our hands the clue to an orderly and natural arrangement of parts; that the structure of the country itself compels the recognition of at least four subdivisions :— ^ I. The coast region south of the Zygos, between the Aspro and the Phidharis, together with the central basin. The similarity in their structure, and the real physical con- nexion which exists between them, justifies the union of the maritime and inland plains under one head. Historically also this union, partially at least, prevailed ; for the western division of the central basin, together with the plains by the sea, constituted the section known as Old, as distinguished from Epiktetos, Aetolia. II. The central watershed. III. The river-system of Agrapha. IV. Eastern Aetolia, i.e. the canton of Kravari, and the basins of the Phidharis and the Mornos. CHAPTER II. The Aetolian Plains. Approaching Aetolia from the sea we have under our eyes the whole extent of the country between the Aspro and the Phidharis. The bare rampart of the Zygos limits the view towards the north. On the east the promontory of Vardssova marks the mouth of the Phidharis. In the west rise the hills of Akarnania and the blue outlines of Kephallenia. Imme- diately before us there is a complex system of lagoons, on the edge of which, four miles away, we distinguish Meso- longhi ^, a name of deathless fame in modern Greek history. To the right and left stretches a narrow broken band of low ground, the 'thread^' that separates the lagoons from the sea. The lagoons are so shallow that only boats of the lightest draught^ can cross them, guiding their course by the piles driven at intervals into the bottom. This was once the only means of reaching Mesolonghi from this side ; but now the embanked causeway which extends right across the lagoon allows direct communication between the steamer and the town. In shape the lagoon* is roughly an equilateral triangle, with a base of about eighteen miles *, and a height of about twelve^. Mesolonghi, the most important town in modern Aetolia, the capital of Aetolia and Akarnania, which together form a single Nomos, stands near the middle point of its ^ Mea-oKSyyi, Me(ro\6yyiov. ' To 'Pa/i/ia. ■^ The ' monoxyla,' as ihey are called. * For a description of the lagoon and the town, see also Trikupis, 'linopia. ii. 361. Reference should be made to the chart on p. 166. ° Between the mouth of the Phidharis and Skropha Point. * From Hdghios Sostis to the head of the inlet, above Aetolikon. lo THE AETOLIAN PLAINS. [ch. eastern side. The name of the town > is evidently descriptive of its situation in the midst of the Ao'yyoj, or ' waste of wood and marsh,' which seems to have covered much of Southern Aetolia after its depopulation by the Romans^. Mesolonghi in these days makes a very different figure from that painted in earlier accounts, or even in those of quite recent date. Seventy years ago Lord Byron wrote ^ : — ' If we are not taken off with the sword, we are like to march off with an ague in this mud-basket ; and to conclude with a very bad pun, to the ear rather than to the eye, better martially, than marsh-ally ; — the situation of Missolonghi is not unknown to you. The dykes of Holland when broken down are the deserts of Arabia for dryness, in comparison.' Now, instead of the wretched lanes and miserable houses ^ that we expect, we find good streets and excellent buildings both public and private. A long and well-built road has been constructed out into the lagoon, serving as an admirable promenade, but otherwise at present of uncertain utiHty. It is true that the smell of the almost stagnant water strikes the visitor as a disagreeable accom- paniment °, but the native assures him that it is harmless. Certainly Mesolonghi is not the fever-hole that one would be incHned to imagine ; although the plain between the Phidharis and the Aspro is extremely unhealthy, the site of the town itself is universally admitted to be an exception. Mesolonghi is a quite modern town.. Its glory is not inherited from a shadowy antiquity; the generation that gave it fame has not yet entirely disappeared. On the north- east of the town is the Hereon, a garden which forms the last resting-place of those who fell in the ' most glorious of all combats, that for their country.' The statue of Byron stands •* -Leake, N. G. i. 113. Melelios, Geogr. ii. 306, says : Meo-oXdyyi, ^ Kpe'iTTov Me(To\dywi'. ' Cf. Strabo, p. 388 ; Koi ^ tS>v AIt(o\S)v Si Koi 'AKnpvavmv iptj/iia k.t,\. ' Feb. 5, 1824. See The Letters and Journals of Lord Byron, Thomas Moore, p. 999. * Cf. Pomardi, Viaggio nella Grecia, p. 34 : ' Le strade sono angustissime, non avendo piu di otto palmi di larghezza, e sono solo lastricate in mezzo con pietre mal connesse. . . . Le case sono basse, e sovente fabbricate con pietre senza calce.' ^ An evil of old standing. In 1766 Chandler complains : ' In the evening the air stunk abominably ; and frogs croaked in chorus without ceasing.' Travels, p. 280. He calls the towns Messa-longia and Nathaligo. ii] THE HEROON. II by the tomb of Markos Botzaris, who was brought from his last field under the shadow of Mount Typhrestos to lie among his fellow heroes. A large tumulus close at hand covers the remains of the unnamed defenders of the town. A few steps away a simple mound is pointed out with a still lingering reverence as the tomb of Byron \ and scattered . under the trees are the graves of many others who played a notable part in the Revolution. As we leave the pleasant shade of the Hereon we feel that Greece has done well to dedicate to her Liberators the soil that drank most deeply of the blood of them and their foes : for the ' Field of Heroes ' is on the line of those famous ramparts which were defended with such brilliant desperation. We may still gaze upon the fosse half-choked with rubbish or filled with green ooze and stagnant water, but the ramparts themselves have almost entirely disappeared: it would have been a miracle had they survived. At one point, however, near the road to Aetoliko, in a line with the outer wall of the Hereon, a few yards of earthwork faced with stone still stand, a relic of the defences that baffled the best troops of Turkey and Egypt ^ From Mesolonghi a long narrow gulf, almost equal in length to the outer lagoon, runs northwards into the interior. About half-way up, six miles north-west of Mesolonghi, lies Anatolikon, or Aetolikon ^, in the mid-channel of the lagoon, which at that point is about a mile in width. A spit of land which runs out into the gulf from its eastern shore, less than half-way between Mesolonghi and Aetoliko, bears the salt- pans of Aliki; north of them AetoHko gives its own name to the lagoon. The town was formerly to be approachfed from either shore by boat alone, but now fine stone bridges and causeways connect the island with the mainland on the east ^ But it is a debatable question. Cf. Gordon, Hist. ii. 117. "^ For an account of the sieges of Mesolonghi cqnsult Finlay, Hist, of Gr. Rev, i. 336; ii. 83: Gordon, do. i. 458; ii. 229': Trikupis, 'lampla t^s 'K\\rjviKf\s 'Eiravaa-Tcia-ems, ii. 361 ; iii. 279. A plan of the fortifications is given by Prokesch-Osten, Geschichte des Abfalls der Griechen. vols, i. and ii. ' 'hvaroKiKov, AiViuXifccSj'. See Meletios, Geogr. ii. 306 : AiVrnXocdi', ^ Koiva- repov ' PivaroKiKov, oitav irpo okLyov xpovov i^fjpx^TO in ttjs Trjs avaKa-)(\a^ov alpa, Kal iyfipopfvov noKKfj ^la nkelov p.ias ir^xeas ! In the popular pronunciation the name of the town becomes Natolik6. Population 5000. 12 THE AETOLIAN PLAINS. [ch. and west. The energy which created these bade fair at one time to raise Aetoliko to a position above that of the capital of the Nome itself; but its site renders the extension of the town impossible, and the debts incurred owing to a too rapid development seem to have permanently crippled its growth. Consequently, although- the streets and buildings of Aetoliko once contrasted favourably with those of Mesolonghi, this is. no longer the case. There is a wide difference between the plain along the coast and that around Aetoliko. From the Phidharis to Mesolonghi the greater part of the interval between the sea and the Zygos is little better than a marsh, covered with reeds which almost conceal the villages from sight. Only as we approach Mesolonghi does this wilderness give place to thriving vineyards and currant-grounds ^. On the north of the town, going towards Aetoliko, we pass through the same uncultivated waste until we reach the more solid ground at the foot of the Zygos, where we again find the currants and grapes of Mesolonghi. It is clear, however, that the soil itself is not to blame for the partial character of the cultiva- tion in this region; the luxuriance of the natural growth testifies to its' fertihty. Want of capital and paucity of population are the two causes that lie at the root of the evil. There is abundant evidence that in Turkish times, to go no farther back, there existed a far more numerous population than at the present day, both here and in many another spot in Greece ; but the effect of the War of Liberation, from the first one of extermination, has been to depopulate the country to such an extent that large tracts have fallen out of cultiva- tion. Nowhere is this more painfully evident than in the plains of Lower Aetolia, where thousands of acres of land admirably adapted for the raising of grain, and gifted by Nature herself with the most convenient means of transport, lie a wilderness without even a single head of live stock upon them. In the territory of Aetoliko, that is to say on both sides of its lagoon, much has been done to repair the evil ; large quantities of currants are grown, but the staple product ' On the east Mesolonghi is surrounded by gardens. The plain to the north appears to have been planted with olive-trees ; they were all cut down during the siege, and have never been replaced, except to the east of the town, and near the Zygos. See Fiedler, Reise. i. 158. II] THE ZYGOS RANGE. 13 is the olive. The groves and vineyards of Aetoliko show what could be achieved if the conditions were more favour- able to the extension of agriculture ; but private credit feels the blighting effect of national bankruptcy, and the system of land tenure is not adapted for the encouragement of individual enterprise. Opposite Aetoliko the western end of the Zygos begins to approach the eastern shore of the gulf; a short distance farther north a narrow passage only is left between the bare rocky foot-hills of the range and the reed-covered edge of the lagoon. Just in the pass are the abundant springs (Kephalovryson), between which and Aetolikon there is a constant stream of animals and carts ; for this is the only good water in the vicinity. Beyond the springs the plain again widens, and forms a large bay towards the east. A conspicuous feature of the Zygos, where it abuts upon the lagoon, is the multitude of huge fissures in the moun- tain. Of these, the most remarkable, and the most curious natural feature to be found in Aetolia, is the cleft opening upon the north-eastern corner of the plain, above the Kephalovryson. This is the Kleisiira ^. Its general direction is from north to south, and it constitutes a natural line of division between the Lower and Upper Zygos '■ The latter is that part of the range which rises above Mesolonghi and stretches eastwards as far as the Phidharis. The Lower Zygos runs west from the Kleisiira towards the Aspro, thus enclosing the head of the Aetoliko lagoon. Extending some way southwards along the western shore of the lagoon there is a succession of irregular hills of no great elevation^. Among them the rocky cone of Haghios Elias is conspicuous from every point in the district. At the end of these ramifications of the Zygos, and exactly opposite the town of Aetoliko, the well-wooded hill Katsa rises from the midst of the olive groves. The space between the aforesaid hills and the Ache- loos is occupied by the plains of Neochori, Guria, Mastru, and Magiila, — the four villages that constitute the modern Derae Paracheloi'tis. These plains extend southwards along ^ 'H KKeiaovpa, 2 Karm-Zuyds. 'ATTUKCo-Zuyds. Cf. Leake, N. G. iii. 552. ' Geologically these hills do not form part of the Zygos. See Map II. 14 THE AETOLIAN PLAINS. [ch. the left bank of the Acheloos, their width gradually increasing, so that, after passing the end of the Katsa ridge, they entirely take up the interval between the Mesolonghi lagoons and the lower reaches of the Acheloos. The general character of the expanse resembles that of the plain on the east of the lagoons. There is the same evidence of fertility, the same lamentable backwardness in the system of agriculture, the same merely partial application of labour and capital to the soil. One remarkable physical feature distinguishes the western from the eastern plain. On the former, rocky heights rise here and there from the waste, like the bare islets that dot the Aegean. The material of which they are composed has nothing in common with the soil at their base ; for the plain consists of a rich yellow clay, unmixed with stones, and in summer reduced to fine powder by the sun. As we look over it from the summit of one of the island- hills it is impossible to resist the cqnclusion that this, plain, like that of Egypt, is the gift of the river, an accretion of the deposits of the Acheloos round the rocky islets at its mouth. In winter the primitive aspect of the scene is to some extent reproduced, when the swollen river rises above its banks, and spreading over the plain once more swirls round the craggy feet of the hills. The pass of the Kleisura ^ forms a natural road through the Zygos into the centre of Afetolia. The beauty of the defile could scarcely be exaggerated. The mountain is riven to its base, and the Upper and Lower Zygos are torn asunder; so as to leave a road between them. We advance from Aetoliko towards the pass as towards a giant gateway leading into the heart of the hills. On either hand the sides of the fissure run up in perpendicular walls, the ledges of which are covered with trees and shrubs. Half-way through the pass the road seems barred, when suddenly it opens in a direction almost at right angles to that in which we have been moving ; then, passing under the now deserted but once very necessary guard-house at the northern entrance, it issues among the woods which cover the slopes of the Zygos on this side. There is a remarkable contrast between the southern and ' Fiedler describes it, Reise. i. 177. Pouqueville's description is a fine example of his method. THE KLKISURA ; LOOKING NORTH. ii] THE TWO LAKES. 15 northern faces of the range. On the former we found only bare steep slopes and grey limestone crags with a scanty covering of stunted bushes ^ How different the picture here ! Forests of chestnut trees and of oaks, gorges choked with vegetation, — the silvan beauties of this landscape have nothing in common with the weary monotony of the view to the south of the mountain. Villages are descried on every side, perched high on the slopes, or straggling along the ravines, or nesthng among the woods at the foot of the hills. The plain into which we gradually descend displays the same contrast ; for it also is well wooded, at least near the Zygos. On its northern edge, beyond the two lakes, the eye rests upon the green hills above Agrinion, leading up, towards the right, to the mass of Mount Viena. Of the two Aetohan lakes lying below us in the plain, the largest and most easterly is that of the Apokuro, called also the lake of Vrachori or Agrinion,— the ancient Lake Tri- chonis. It is a splendid sheet of water, forming a crescent along the Zygos, twelve miles in length and three in breadth. At its eastern extremity, where the mountains round Petro- chori fall almost sheer into the water, its depth is popularly believed to be unfathomable. Even in the calmest days the lake dashes upon the rocks with the fury of a miniature sea, and for this reason, perhaps, it has gained the name U.ikayos. Fish in great numbers and of many kinds ^ are found in it, but few are caught. It is the rarest of all sights to see upon its broad bosom anything in the shape of a boat ; only at Kainiirio on its northern shore can a ferryman with difficulty be found. The lake of Anghelokastron, or, as it is often called, the lake of Zyg6s ^ has a circumference about one fourth of that of Vrachori, and its shores are almost entirely marshy ; for, while the lake of Vrachori occupies almost the entire space between Mount Viena and the Zygos, that of ' That is the character of the limestone mass, north of Mesolonghi, which forms the western end of the mountain ; further east, in the direc- tion of the Phidharis, where we are again upon the sandstone, there is the same aridity and nakedness, perhaps due to the salt breezes from the sea. ^ Strosidhia, Glanldhia, Dhromitsais, Tseriichlais, Chelia, Kephalik^, Theringhia, Biilkais, are a few of them. " For the genesis of the names ' lake of the Apokuro,' ' lake of Zygos,' see p. 228. i6 THE AETOLIAN PLAINS. [ch. Anghelokastron has no high ground near it, except the spurs of the Zygos on its southern shore. Consequently, the western lake makes a very poor appearance as compared with its neighbour on the east. In order to reach Agrinion from the Kleisura we have a choice of routes. We may turn to the west and make the circuit of the lake of Anghelokastron, or we may go eastwards round the lake of the Apokuro. The third and most direct route lies across the morass that intervenes between the two sheets of water. Some two centuries ago the modern main road was anticipated by a certain Ali, Bey of Vrachori, who built the old causeway^ which extends two thousand yards across the treacherous marsh a short distance east of the present highway. If we are to believe Pouqueville and Leake, the Turkish engineer made use of the foundations of an earlier structure, dating perhaps from Roman times. This is not impossible : but none of the earlier work is now visible. Not only the foundations, but to a great extent the Turkish piers themselves have settled in the bog until the water reaches the keystone of the arches., Nor, again, do there remain any traces of the rings for mooring the barges engaged in the commerce of the villages round the lake that are mentioned by Pouqueville ^. The number of arches in the causeway is supposed to be three hundred and sixty-six. Though the old viaduct is now used only as a short cut, and is allowed to fall year by year into ever worse repair, it is a pleasant and worthy continuation of the Kleisura. We ride through the cool shade of oaks, planes, and wild olives, which are festooned with wild vines and creepers, and blocked with tall reeds so as to make an impenetrable jungle. The songs of birds add that charm which is so often painfully wanting in the Greek woodlands. A gentle stream flows through the arches from the eastern into the western lake, and finally into the Acheloos near Anghelokastron. In one hour from the causeway of Ali-bey we reach the bed of the river or torrent Eremitsas^, flowing from the Arabokephalon mountains in the north-east into the lake of Anghelokastron. From the Eremitsas it is one hour more to Vrachori*, officially called Agrinion; the second town ' 'H yf(f)vpa Tov 'AXij^ittctj, or 'AXa;i7re)j. " Vqy. iii. 515. ' 'O X"Vap/30S 'EprinrjTiTrjS. See p. 184. * BpaxSipi, 'E^paioxupiov. ^HfcWf^f , }'■ ?K ■ '■1 1^8 ^^^^^Bl^^^fl^ v"A CP'^^i ^^^^^^^^^^^^^^^^^^^^^^^^■e ''' .,-1:*'.. - i-fMjV ^ -^ Y7^- /'i-C'3- £ '-^ -1 a?^vin:^«iM> ° ' ^^' if -i ■ JT^* • --^'l 4\ IB J?ii.J>,;y " '^'Sl j^KKr " ,t'; ;,r" 'M w^ rfi wm i ^**^« ■|K»" Vi ^' ■■ ' Fr f^W, 4 i, fi-i m:-;' .f ■ 'J ■sM k:^ll '•^■'■ft**"' 'i&t!^WgB^ i;Rt\k ■t #^ f^ p'-y i"®i' ■' Mf» ^Hk W^' ^^^p."|;;^JI KI- •::'■ >?■'., l^.t-.l fl^w )!gg,|J iKt "^i^ t \'", ■W^Wi^ W^P jK ' w \'M 'S^^^^^^mJiB-'^f ^mbBIp- WBk ' "^p ''jI ^^^^^K-^W ■ Ij !*■■'"•*-' ^^^^^^^^^^^^^^^^K; ^'s'^^^pK 1 ill ^^^^^K^" ,1 J| ri' ' i^jpjg '' ^ « <^ ^^'-; ^n^p' r' .^ Rvi ^^m^tf Ki'^-v 3 -1 iiefiiiiii H P O tfl a g, S o o ►J o X u < o II] APPEARANCE OF VRACHORI. 17 in Aetolia. The word Vrachori seems to be a corruption of Hevraiochorion, the ' town of the Jews.' At the time of the Epanastasis there was a considerable Jewish colony here, which fared very badly ^, Agrinion is a beautiful place. The site is well chosen, on the long slopes at the foot of the Lykorikia^, the verdant ' Wolves' ridge ' which stretches along the northern edge of the plain. It is thus sufficiently removed from the unhealthy atmosphere of the western lake and its attendant pests, yet not in so lofty a situation as to render com- munication with the plain difficult. The tobacco-fields, vine- yards, and fruit-gardens, dispersed among the houses, give the place the appearance of a group of villages rather than of a town. Great imf)rovements have been carried out in Vrachori during the last few years, so that the guide- book description of its narrow ill-paved streets is no Ipnger true. It is the present terminus of the Railway of Northr West Greece^, and numbers over seven thousand inha- bitants. The only drawback under which it labours is the scarcity of water. To prevent the waste consequent upon employing the drinking-water for the purposes of irrigation, the idea has been entertained of supplying the gardens by means of a canal from the Acheloos, some six miles distant ; but, the funds of the Deme having been expended in the construction of a useless and ugly Platia and on other im- provements, this scheme remains still in abeyance. The charm of Agrinion is certainly to some extent acci- dental and dependent upon the oasis-like character of its immediate surroundings. There is little to attract in the contemplation of the western section of the central Aetolian basin. North of the lake of Zygos the plain stretches to the hills beyond Stratos, for the most part a desolate and uncul- tivated waste. In the vicinity of the lake itself the country is a swamp, covered with reeds and the remnants of the vast forest that once flourished here. As we advance north- wards the reeds and trees disappear, and we enter upon a bare expanse which, although one of the most fertile in Aetolia, lies but half-cultivated, like the plains of the Para- cheloitis. The five villages in this section contain an ' Cf. Finlay, i. 202. " Auxoppdicta. ' SiSrjpdSpo/ios TTJs Bopcio-AvTiKrjs 'EXXaSor. C i8 THE AETOLIAN PLAINS. Lch. aggregate population of less than two thousand souls. It was not always thus. Meletios, for example, in the seven- teenth century, mentions one of these villages, Zapandion, two miles west of Agrinion, as being one of the most important places in Aetolia ^. In the rising of 1821 the six hundred Turks inhabiting Zapandi maintained themselves for more than a month against great odds, but were finally overcome and put to the sword 2- They have left their traces in the two ruined minarets that stand conspicuously on the outskirts of the village, the only examples now exist- ing in Aetolia proper. When Bazin visited the site in i860 he found only a mass of ruins, among which were a few wretched cabins^. In 1893 Zapandion once more exists, a village of three hundred inhabitants, much given to the cultivation of tobacco. The history of this place serves to call attention to the undoubted gradual increase in the population of Greece, but the upward tendency is hampered by the economic difficulties of the country. The shores of the eastern lake present a different spec- tacle*. Numerous villages occupy the plateaux above the narrow but fruitful space on its borders, and the rich crops of maize and tobacco gladden the eye with their story of prosperity and comparative wealth. And the ruins bear witness to an equally flourishing state of things in ancient times : scarcely a village is found here to-day that does not mark the site of an Hellenic town. The road lately com- pleted round the lake makes it easy to visit this most important part of Aetolia. Both for the archaeologist and for the ordinary tourist the circuit is one of supreme interest. The thriving and hospitable villages planted at moderate and regular distances solve the chief difficulty of Aetolian travel, and allow us without anxiety to dwell upon the natural beauties that surround us. Our road lies through maize- fields and vineyards, beyond which the lake glistens in the ' Geogr. ii. 306 : KmnomSKcis Zanavn, BpaxSipi, MetroXdyyi. ^ Cf. Fin. i. 203. Gord. i. 261. '' Mem. p. 275. * Cf. Oberhummer, who says the lake of Apokuro ' bietet durch seine von bewaldeten Bergen eingeschlossene Lage einen der landschaftlich interessantesten Teile Griechenlands, indem sich dort der Charakter unserer Hochlandseen mit dem Reiz sudlicher Farbung und mittellandi- scher Vegetation vermischt.' Jahresbericht der Geogr. Gesell. Miin. 1885, Heft X. n] THE EASTERN SECTION OF THE BASIN. 19 distance, or over the rocky spurs which thrust themselves through the fertile belt into the blue waters. The prospect changes from hour to hour. Full of character and variety, the mountains give to every scene a background of which we never weary. Blessed as it is with a pure and healthy air, a fertile soil, a perennial water-supply, and fenced round with mountain barriers, we feel that this region three and twenty centuries ago must have been the real heart of the Aetolian land, and the home of its most powerful chieftains. c 2. CHAPTER III. The Great Watershed. The beautiful grey pyramid of Mount Viena \ which is so conspicuous from every point in the central plain of Aetolia, warns us of the character of the region that lies to the north of Agrinion between the Aspro and the head waters of the Phidharis. It is one of the wildest in Aetolia. We bid adieu to the plains as soon as we enter it ; for from Agrinion northwards, until we burst through the Pindos upon the valley of the Spercheios or the great level expanse of Thessaly, there is nothing that deserves to be called a plain. All is mountain, a bewildering and savage land of peaks and ravines that defy description. The roads are narrow and often dangerous tracks, along dizzy slopes above the gorges. The rains and snows of winter disintegrate the rocks, and when the grip of the frost is loosened landslips carry away or bury the path, and compel us to difficult detours. As we mount ever higher we turn to look upon the panorama of the central basin. The humps of Rhigani, Vardssova, and Klokova, peep over the long green Zygos, and the dim blue forms of the Peloponnesian mountains rise on the southern horizon. The habitations and works of men are left behind; wandering Vlachi seem to be the only occupants of these wilds. Here and there only, at the bottom of some gorge, a few scattered houses are descried, as of men who have exiled themselves from the world to the recesses of the mountains. Such are the villages of Stregania and Lambiri^, which, buried among the masses of the Arabo- kephala '\ we pass on our way to the north. It is with ' BijeVa, a contraction from (Kvpla) Eiyeveia. See Leake, N. G. i. 131. " 2rpi)yai/i(i. Aa/impiov. ' 'Apa|3oKe0aXoi', 'Apa0o(ce0a\a= Negro-heads. THE MONASTERY OF PROSSOS. 21 a feeling of relief that, having crossed the water-parting, we look down at last upon a large village lying on the western slope of the valley before us. This is the village of Prossos ^ ; it contains about twelve hundred inhabitants. The houses rise in terraces above the torrent to which the village gives its own name. Vines, maize, walnut and mul- berry trees, grow in rich profusion about them. It takes us half an hour to thread our way among the terraces to the pass at the opposite end of the valley, where at the foot of great precipices we see the famous monastery of Our Lady of Prossos^- It is disappointing to find that the monastery has so modern and poor an appearance. The foundation dates from the ninth century ; but the ancient buildings were fired by the Turks, and with them perished the manuscripts and Deeds , of the monastery. The Panaghia, who has wrought many other miracles at this favourite shrine, pre- served from the infidel flames the inner and most holy portion of the church. The sacred edifice, being really double, is built so close to the cliff" that the small dark inner chapel, containing the ancient Eik6n of the Virgin, is merely a cave in the rock ; and when the body of the church in front perished this part remained intact, protected from the Turks by the flames which their own hands had kindled. It is a privilege to be admitted to the cave in order to gaze upon this venerable Eikon, painted, we are assured, by Saint Luke himself Brass and silver repousse work conceal all the painting except the face and hands of the Mother and Child. It is dark with age, and is indeed in wonderful con- trast with the Eikon of the outer church^. Tradition says that it was brought from Brusa, near Constantinople, whence came also the name of the village. The favour of the Virgin still attends the monastery. The monks, six or eight in number, point out a large hole in the rock above the church, and tell how a huge mass falling one day from it should, by all the laws of nature, have ground the building to powder, but leaped instead over the outer wall of the monastery and crashed into the torrent. The great annual Pan6gyris, or pilgrimage to the shrine, is attended by large crowds, in spite ^ Ttpoa-aros, Upovaos. ^ To novacrrfipi rrjs Ilavayias Tov Upo(rcrov. ' As to the value of the picture as a specimen of Byzantine art I know nothing. 22 THE GREAT WATERSHED. Lch. of the difficulty of reaching the place. And not seldom miraculous cure rewards the afflicted worshipper for the piety and faith that have brought him so far to kiss the holy Eikon. When we mount to the path along the precipices above the monastery we realise the grandeur of the situation. A narrow paved track along the edge of the cliff, several hundred feet above the serpentine torrent of Prossos, is the only exit from the valley. At our feet, so close that it needs but a single step to hurl us into the court below, Hes the monastery. The vale of Prossos, running into the bosom of Arabokephalon, constitutes the prospect towards the south and the quarter from which we have come. Northwards, we gaze along the course of the torrent to where the giant Kaliakudha rises on its eastern bank, fronting the scarcely less huge Chelidhoni ^ on the west, — a mighty portal through which we pass to Karpenision. Along the eastern side of the torrent rise precipitous slopes, similar to those on which we journey, but still in the shade long after the sun has flooded our own path with light. On the dangerous tracks that run like fine threads along these opposite slopes we see women creep slowly, tiny as ants, under a golden burden of grain from the scant patches that gild the mountain side : no horse could keep its footing on the perilous shingle. At length the Prossos torrent which we are following joins the stream of Krikellu coming from the east round the southern foot of Kaliakudha. This warns us that we have reached the base of the great watershed of Aetolia: beyond this point the rivers and mountains run no longer from east to west, but from north to south. If we look at the map we see that it indicates only five large villages on this watershed. Prossos is almost on the line bisecting the region from north to south. On the east, along the upper Phidharis, are the villages of Koniska, Ara- chova, and Klepa ^, of which only the first numbers over one thousand inhabitants. On the west of the central line the only considerable village is Haghios Vlasis *, with less than one thousand inhabitants. The eastern section does not differ in nature from Kravari, ' KaXiuKoCSa. XeXidrai/i. ''■ Kai/KTKa. 'Apaxo^a. KXeirS. ' "Aytos B\d(TtjS, 'A. BXao-tri;;. '"] THE WESTERN SECTION. 23 which canton begins on the southern bank of the Phidharis. The villages named are all on the southern side of the ridge interposed between that river and the stream of Krikellu; their connexion is naturally with the canton of Kravari rather than with the rude and unproductive region that lies to the north. The western section, lying between Prossos and the Acheloos, bears an entirely different character, and one that has little in common with what we have hitherto seen. We remark in the first place that the range of Arabokephalon does not follow the direct line to the west, but breaks to the north and north-west, being continued in that direction by Plokopari, Kynigu, and Kutupas^. The striking serrate ridge formed by those three mountains is, in fact, the only natural feature of importance in the whole of this western district. The country is savage, but with a savagery very different from the grandeur of Prossos. The lofty peaks give place to a net-work of hills, for the most part of very moderate elevation. Even the vegetation is changed, and the heights are covered mostly with brushwood, — purnaria, arbutus, and philyki^, intermixed with bracken. We no longer toil along dangerous paths cut in the rock, but wind through a labyrinth of hills on crumbling tracks ; we have in truth reached a different geological formation, — no longer hard limestone, but soft sandstone, which splits into regular cubes and finally crumbles to gravel. The bodily fatigue attendant upon travel in the eastern portion of the region is replaced by weariness of spirit ; for the monotony of the hill-forms and their utter want of character make the road tedious. The prospect changes without real variation. Only along the banks of the rare streams can we find relief from the ennui of this desert : by no other word can this region be described, in spite of its green shrub-clad hills contrasting so strongly with the naked peaks of the limestone formation. The sameness and meanness of the surroundings render more conspicuous the ridge of Kynigu and Kiitupas in the centre of. this sad region. This remarkable series of sharp ' nXoKOTrapi. Kuj/TjyoO. KouroDTraf. ' Tlovpvapi, npivapi, mpvapi, — Quercus coccifera ; of which there are several varieties. Kovnapid, Arbutus unedo, L. iikvKri, Phillyrea latifolia, and Ph. media, L. 24 THE GREAT WATERSHED. peaks, resembling the teeth of a huge saw, is visible for miles in any direction. Travelling along the high road by the shores of Lake Vrachori, we catch a glimpse of it in the north beyond Mount Viena ; and it is not until we are buried in the heart of the mountains of Agrapha that the famihar jagged line disappears from our southern horizon. On the western slopes of Kutupas, at an elevation of 883 metres above the sea level, lies Haghios Vlasis, the capital of the Deme ^ The majority of its nine hundred inhabitants live scattered in dependent hamlets, or Machaladhes. It would be difficult to find in Aetolia a site with a more splendid and extensive prospect. In the foreground are the low bushy hills, so characteristic of the district, among which we trace the course of the Acheloos for many miles from the north. Beyond them we have the panorama of the Akarnanian mountains purpled with the glory of the sun as he sinks to the Ionian sea. During the summer the principal families go down to the Acheloos, where, some three hours from Haghios Vlasis, are the hot springs of Kremasta ^. The springs rise in the bed of the river itself, near both banks, and, as they are considered especially efficacious in cases of rheumatism, large numbers from all parts of Aetolia visit them annually. They contain iron and sulphur, like the hot springs near the village of Murstianu ^, on the southern shore of the lake of Zygos. The visitors erect huts for themselves of branches and planks, and take with them provisions to last during their stay. ' TSi> XLapaKajiTrvKlav. " 's Ta KpefiaiTTd. For the meaning of the name see Leake, N. ,G. iv. 253 : ' Not far below Tripotamo, the river is said to flow between preci- pices so closely approaching as to be crossed by a bridge of ropes, whence the place is called Sta Kremasta.' ' Movp^ndvov. The accommodation at this place is of a more comfortable character, and a small village rises round the springs during the season (June and July). They lie exactly opposite Agrinion, about one hour and a half east of AnghelOkastron. o o a o < M 33 H h O O K o o CHAPTER IV. Agrapha ^. Looking from some elevated point in the region just described, say from the peak above Haghios Vlasis, we see at once the structural contrast between the country on the south and that which remains to be surveyed on the north. iTo the south and east we trace the line of the central water- shed, of which our standpoint is the last member to the west. Beyond it rises the Zygos, and beyond that are the moun- tains of Achaia. In the depressions lie the central plains of Aetolia, and the lowlands of the coast. All these features run from east to west, in lines nearly parallel to each other. Turning northwards we trace, on the extreme left, the great gorge of the Acheloos ; and as our gaze travels slowly east- wards it rests in succession upon the mountain rampart of Pteri above Granitsa, the river and mountain-chain to both of which Agrapha gives its name, the Megdhova river ^ coming from the confines of Turkey, Chelidhoni and the mountains that lie between the Megdhova and the Karpe- nisiotikos : lastly, Veluchi closes the view to the north-west. We have before us a river and mountain system of which the members run in practically parallel lines from north to south. The Agalianos'^ river at our feet, flowing from south- east to north-west, is the great base-line upon which these natural perpendiculars are raised. Were it not for the sudden break northwards made by the ridge of Kutupas, the course of the Agalianos would not have been compelled to > Ta* Ay pa(fia. " MeySo^as. ' 'AyaKmvos. So called from the village of Agalian6s, on its southern bank, nearly opposite the point at which it receives the waters of the Megdhova. 26 AGRAPHA. [ch. that bend towards the north-west which mars the symmetry of the physical geography. A further significant relation between the mountains and the streams must be observed. Although it is sufficiently near the truth to speak of them as forming parallel lines from north to south, yet as a matter of fact the courses of the rivers are somewhat oblique to the mountain lines. The Agraphiotikos and the Megdhova trend to the south-west, whereas the interposed ranges of Pteri and the Agrapha mountains lie due north and south ; and this fact has important bearings upon the character of the region \ As we study from our lofty standpoint the broad lines of the landscape, the eye anticipates much that we shall after- wards realize in travelling through the district. Rivers of large size for Greece, mountains overtopping any seen since we looked upon Vardhusi, small and widely scattered villages, scanty evidences of cultivation, — all warn us that the difficulties thus far experienced are likely to re-appear in an Aggravated form. There are three routes by which we may penetrate this region from the south. We may follow the road by Haghios Vlasis on the west of Mount Kiitupas, thus traversing the western sandstone belt. An alternative route is that which crosses the limestone Arabokephalon mountains, by way of Prossos. Or, if our point of departure lies in the east of Aetolia, we may keep to the eastern sandstone region, and by skirting the western side of the Oxya range fall into the Karpenisi valley by way of Krikellu. These three are the only routes leading northwards from Southern Aetolia ; all others that may be indicated on the maps are either spurious variations, falling ultimately into one or other of the three main passes, or mere mountain tracks, impassible to a loaded animal, and so without any claim to be recognized as routes, of communication. Of the three routes, that by Haghios Vlasis is by far the easiest, but it is not the shortest line of communication between Central Aetolia and Karpenisi, the northern capital. The pass of Prossos, though by much the most difficult, is the direct route, and the artery of trade between North and South Aetolia, so far as trade can be said to exist. Goods ' As regards, for example, the means of communication. iv] THE PASS OF PROSSOS. 27 landed at Mesolonghi and taken by rail to Agrinion, or brought by caravan from Karvassaras ^ on the Akarnanian coast to the same place, are transported thence by the mules of the Karanghunidhes ^ of Stratos through the pass of Prossos in three days to Karpenisi. Commodities from the east are landed at Stylidha, the port of Lamia, and are carried without difficulty up the valley of the Spercheios, which extends to the confines of Aetolia. Beyond Karpenisi, to the north and north-west, there is nothing but the wildest mountain country, the passes of which are closed during the winter months. In our survey of the central watershed we retraced our steps from the junction of the Prossos torrent with the river of Krikellu. It is a remarkable point, a striking combination of mountain and river. The united streams flow in a north- westerly direction, round the base of the enormous mass of Kaliakudha, until after a short distance they meet the river of Karpenisi pushing its way from the north-east between the same mountain and Chelidhoni. Mount Cheli- dhoni and the ridge along which we have travelled from the Prossos monastery are separated by a deep narrow gorge, through which the three streams roll towards the north-west. The confluence of the Karpenisiotikos with the combined Prossos and Krikellu streams from the south takes place at the very mouth of the gorge, where it is spanned by a bridge, built in 1892 but 'swept away by the floods of the same winter^ The path from the bridge northwards is as wild and grand as that from the monastery. The Karpenisi river runs in a tortuous course far below us; Kaliakudha and Chelidhoni thrust forward alternately a craggy foot to crush the stream, which writhes like a captive snake in its effort to be free. Farther along it seems as if the mightiest forces • Kapfiaa-ixapas, derived from the Turkish ' Karavanserai.' It is the ancient Limnaia, at the southern end of the pass of Makrynoros, the Thermopylai of Western Greece. == KapayKoiviSes, 'Black Cloaks,' the name given to the Akarnanian Wallachians. See Heuzey, Le Mont Olympe. p. 267. ^ In 1893 it was in course of construction for the second time. Yet there are no engineering difficulties to be overcome, as the cliffs on each side serve as piers : it is simply a matter of workmanship. For the similar state of things in Turkey see Ramsay, Hist. Geogr. of A.M. p. 82. 28 AGRAPHA. [cH. of Nature had been called into action to rend a path for the torrent. Two tremendous red precipices formed by the two mountains approach each other so closely that originally room for the stream alone was left; but now a narrow path for horse and man has been hollowed along their sides. On each bank, at the narrowest point of the gorge, there stands a small shrine ^ A short distance beyond this place the mountains again approach each other, and this time form a ravine impassible to all but the river itself. All exit from the valley is closed, except that by a path which climbs the steep spur of Chelidhoni. A painful effort of three quarters of an hour brings us to the top of the ridge, — our goal is in sight, for there, seven miles up the narrow valley to the north, Veliichi rears his bald head, and we descry the houses of Karpenisi^ nestling at his foot. Our guides, as they reach the chapel on the ridge, cross them- selves devoutly as if in gratitude for their escape from the long and dangerous pass. From this point the road is easy ; first to the ' Little village^,' opposite which, on the other side of the river, the two hundred red-tiled roofs of the 'Great village*' contrast brilliantly with the dark forests of Kahakiidha. Then we descend into the narrow valley down which flows the strag- gling Karpenisi river; the valley is in fact almost entirely converted into a channel for the stream by the network of canals that distribute the water among the maize-fields. The path is bad, of mud or cobbles, generally of the two combined, and along it we splash for nearly three' hours towards the foot of Mount Typhrestos. On either hand, rounded thickly wooded hills confine the valley; they are ramifications of Kaliakiidha and Chelidhoni, ending towards Karpenisi, — on the east in Koniska^, with the village of Klavsion at its foot; on the west in the hill and village of Koryschadhes. As we advance, the main peak of Veluchi gradually disappears behind the lower heights of the mountain, so that by the time Karpenisi is reached the ' Konisma ; that is, e'lKonvuriia. " BeXoup^i. KapTrci/^ciov. ' M.iKpoxiopi6, for MiKpoxaipwv. * MeyaKoxtopw, or Tpai/o^Spi. But Neumayr {Denkschr. xl.) is wrong in saying that in Aetolia rpavos is always used instead of fuyoKos. ^ Kavia-Ka, from its shape. THE PASS OF PROSSOS: LOOKING NORTH UP THE KARPENISI RIVER. IV] THE VALLEY OF KARPENISL 29 summit is quite invisible. Hence Veliichi is somewhat dis- appointing. It is only from a comJ)aratively great distance that the peak can be seen rising triumphantly over the excrescences at its base. As beheld in the early spring from the Maliac gulf, before the snows have melted, the mountain is a splendid object on the western horizon. From the foot of Koniska the valley runs up in a long slope towards Veluchi, and also extends, about a mile in width, eastwards until it ends at the ridge that connects Veluchi with Oxya and Oita. At the corner, where the valley bends eastwards, a small conical height is left islanded to the south of the town; it is crowned with the church of Haghios Dhimitrios. The Karpenisi river follows the configuration of the valley, so that, after rising above the village of Laspi, it flows first to the west and then bends sharply to the south along the base of Koniska. As we climb the aforesaid long slope we find ourselves scrambling over the dry stony bed of a torrent descending from the broken hills on the west, which bear the name Misorakia; this torrent unites with a second similar torrent of cobbles coming from the bosom of Veluchi itself and dividing the town in front of us. There is, in fact, no other entrance to Karpenisi from the south than that constituted by this dry gully, from which we emerge into the narrow, steep, rudely paved lanes of the town. The houses climb the hill on both sides of the torrent in irregular terraces. The tortuous main street, after cross- ing the torrent bed, issues on the eastern side of the town as a broad carriage-road^, cut in the hill above the pro- longation of the valley as already described. About half a mile from the town, owing to the gradual fall of the road and the steady eastward rise of the valley, we find ourselves actually crossing the level in a south-easterly direction. Here, under the grey peak cut with the severe beauty of a Greek gem against the transparent blue, is held the great annual Panegyris ^ of Karpenisi, during the first three days ^ Along which there is regular communication with Lamia and Stylidha, and so with Athens. 2 navrtyvpis iforopiKri ; as opposed to the n. iKK\r)aia o iv] MARKOS BOTZARIS. 31 It was in August, 1823 \ that Mustals, pacha of Skodra, was leading the Gheg ^ Albanians through Agrapha to effect a junction before the walls of Mesolonghi with the Tosks * of Omer Vrionis. Mesolonghi was the only town in Western Greece that still held out; the Aetolian and Akarnanian chieftains were too busily occupied with their own quarrels to spare time for engaging the common enemy ; the Ottoman fleet threatened the coast from Kandhih to Naupaktos; the leaders in Agrapha and on the Upper Acheloos had fled or submitted. All Western Hellas seemed lost. The Turkish advanced-guard commanded by Djelaleddin Bey, the nephew of Mustai's, encamped four thousand strong round the spring at the foot of Koniska. The three hundred and fifty Su- liotes of Botzaris had at last been joined by Karaiskakis, the Tsavellai, and others, so that the united forces of the Greeks numbered about twelve hundred men. Markos encamped in Mikrochorio, the other chiefs in Megalo- chorio on the left bank of the river. It was impossible to meet the Turks in the open field ; it seemed likely that the national forces would after all disperse without a battle. Botzaris succeeded in inspiring his companions with his own brave spirit, and a night attack was planned. Some of the Suliotes had entered the hostile camp and reported its dis- position ; being Albanians they were able to do this without fear of detection. Five hours after sunset Botzaris was to attack from the valley, and the other leaders were to support him from the hills. The Suliotes kept their word. A quarter of an hour after the appointed moment their war-cry startled the sleeping Ghegs : but their attack was unsupported. Envy and fear paralysed the arms of the Greek captains and their men, and only Kitsos Tsavellas with his brother and a few others came down to share the peril and glory of Botzaris. The heroic Suliote was wounded, but he pressed forward to the low wall surrounding the tents of Djelaleddin and his staff. The veteran Ghegs were as familiar as their enemies with nocturnal warfare, and were trained like them to fire with ' Cf. Fin. ii. 10. Gord. ii. 32. Trik. iii. 62. For an example of the transformation of history into childish fable, see Pouqueville, Histoire de la Regeneration de la Grece, iv. 5 fol. 2 IV/kOI. ° ToCTKlbfS. 32 A GRAPH A. [CH. deadly precision where any but Albanian eyes would have- been at fault. The head of Botzaris, raised rapidly above the wall to discover an entrance, was outlined for an instant against the dusky sky ; a ball sped to his brain and he fell dead. As dawn approached, the Suliotes became aware of their loss. His cousin Athanasios Dusas took the body of Markos upon his shoulders, and with an immense booty of arms and horses the force retired unmolested to Mikro- chorio. Thirty-six men were lost in the attack. The richly ornamented weapons won that night marked out for long afterwards the Suliotes who shared in the exploit at Kar- penisi, but they felt that their splendour had been dearly purchased, and all Greece mourned the loss of the gallant Botzaris. His body was brought down to Mesolonghi and buried in the Hereon, but it was not until quite recently ^ that a memorial was raised on the spot where he died. Tsaveilas, Kara'iskakis, and the other Greek chiefs, attempted to hold the pass between Kaliakudha and Chelidhoni, but they were defeated and driven from a position that five hundred resolute men could hold against the world. It is worth the time to make the ascent of Veliichi for the purpose of gaining a clear idea of the geographical relationship between Aetolia and Thessaly. No other Aetolian mountain is so admirably adapted for the purpose, and perhaps none is so easy to climb. The path leaves the town by the eastern side of the gully which we have remarked as descending through its midst ; then, following the aqueduct from the Rovia, we turn to the left at the head of that depres- sion and reach a belt of firs, most of which have had their heads lopped by the shepherds. Above these comes a grassy but otherwise bare plateau, called Rovolakka, at the base of a steep stony hill bearing the appropriate name of Sa'itani ^. Surmounting it we find ourselves looking down upon the plain on which the Panegyris is held ; immediately before us is the narrow ridge called Samari, the ' saddle ' between the revma Sostrunka on the left and the terrible chasm on the right extending to the level of the plain. At the head of the chasm the wind thunders in the cliffs just under the summit of the mountain. These cliffs are called 's ta Ghid- hia ■■', as wild goats are occasionally found in them. The hill ' 1893. 2 SalVni'i = Satan. ' 'sraTi&ia. IV] ASCENT OF VELUCHI. 33 beyond the saddle is extremely steep and stony ; its summit is a level platform dotted with a few rocky knolls, under one of which a shepherd's enclosure provides the traveller with shelter for the night. The main summit of the mountain is then distant about an hour's climb. An intensely cold spring gushes forth close to the top. Greek springs are partial to these lofty situations, being often found on the mountain when the plain around is parched and dry. Peirene on the Akrokorinthos and Hippokrene in the bosom of Helikon are classical examples ^. It is possible that a sudden change in the wind may enshroud the peak and these higher parts of the mountain in a dense fog, often lasting several days, and this, combined with the north wind rushing from the Thessalian plains, makes the situation anything but a pleasant one. The ascent may have to be made several times before the view of all the world, as the Karpenisiotes call it, can be obtained ^. If the state of the atmosphere be favourable, the Aegean can be seen on the one side and the Adriatic on the other. In the space between Veluchi and the Aspro there are only three villages of any importance, — Agrapha, Kerasovon, and Granitsa. Kerasovon lies to the north-west of Kar- penisi, at nearly the middle point of a line connecting Karpenisi with Granitsa. Agrapha is almost due north of Kerasovon, so that the three towns are situated on the angles of a nearly equilateral triangle. As regards position, Kerasovon^ is the most important village of the triad, lying as it does on the southern face of a peak 1760 metres high, the last member of the chain of mountains interposed between the Megdhova river on the east and the Agraphiotikos on the west, and extending northwards beyond Agrapha. This range is in fact to be regarded as the true continuation of the Pindos*. Kera- sovon thus commands the defiles along the banks of the two rivers ; in other words, the roads to the north, for in ' A spring rises on Mount Kiona, one hour from the summit. ^ It is the general belief that the lights of Constantinople can be seen from the top. But this is said of any and every peak. The Plataniotes have the same fable of the hill above their village, though that is only 1400 metres high ! » Kepda-o^ov. * See p. 44. D ipi Trjs TaTapvas. H < w X H s o ipi ToS KopiKov, or KopaK.'ou. Leake, N. G. iv. 269, calls it the bridge of St. Bessarion, built by the monastery of Dusikon in Kotziaka. 38 AGRAPHA. [ch. from Kardhitsa and Trikkala in Thessaly converge upon the Acheloos in their course through northern Agrapha to Arta and Epiros. Between these two roads leading across the Acheloos, the one by the bridge of Tatarna, the other by that of Korakos, there is interposed the mountain complexus of Agrapha, an almost insurmountable barrier to trade and communication. Buried as it were at the bottom of a basin, in the heart of this wild land, is the village that bears the name of the district,— Agrapha, a word which in North Aetolia is synony- mous with difficulty and danger. Wild and difficult passes, the gorges of the rivers, lead into the narrow valley where, perched on a ledge high above the stream, the few houses of the village cling to the mountain as if ever in danger of sliding down the slope and dashing over the edge of the precipice. Nor is the danger purely fanciful. In 1877, after long-continued rains, the whole western half of the village collapsed, and began to slide steadily downwards, finally sinking into the quivering earth ^. In Agrapha the Demarch's house, sadly out of the perpen- dicular, alone survives as a specimen of the older architecture which came to this untimely end. The four plain walls, pierced with a single doorway and furnished with a few loopholes to admit air and light to the rooms on the ground floor, show plainly that the original idea was that of a fortress rather than a dwelling. These lower chambers are used as store-rooms, and sometimes for cattle. Mounting the steep dark staircase in the inside we emerge into a long hall running the entire length of the building, and taking up half the breadth. At each end of the hall there is a platform on which the master sits, which at night serves as his own bed-place and that of the other male members and ordinary guests of the family. The walls are decorated with guns and pistols of curious clumsy-looking shapes, richly inlaid with silver, to most of which there attaches some grim legend of fierce and bloody encounter with the Turk. Door- ways lead into separate smaller apartments, such as kitchen and sleeping-rooms for women and distinguished visitors. ' Cf. the fate of the village Lykochori on the slopes of Viena ; ' it fell down the side of the mountain and disappeared with all the adjoining soil.' Leake, N. G. i. 131. iv] THE DEMARCH'S HOUSE. 39 Numerous openings in its three outer walls admit light and air to the great chamber through carved wooden lattices, the handiwork of the Wallachians inhabiting the range of Pindos. In summer the room is pleasant and cool, but in winter, when snow blocks the passes and fierce blasts tear howling through the gorges, it must be terribly cold. The hall projects at the ends, and often on all three sides, beyond the line of the outer walls, so that it forms a sort of closed balcony, the ends of the supporting beams being in many cases richly carved. Such is the house of the Demarch of Agrapha, certainly the finest and perhaps the only example of this type to be found now in Aetolia, though many still survive in Thessaly, especially in the villages on the slopes of Pindos and Pelion. This style of architecture seems to be due to Wallachian influence, for apparently it is found only in those parts of Greece which were once in the hands of that people. It is well known that Thessaly formed the kingdom of Great Wallachia, while Little Wallachia embraced part of Aetolia and Akarnania '^. In essentials, indeed, the building we have described is identical with the type prevailing through the length and breadth of Greece, but the Greek has apparently quite lost that artistic instinct which might give individuality and interest to his dwelling if it failed to give it beauty. The Wallachians seem to be the sole repositories of art in the kingdom ; as workers in wood or in silver, as embroiderers of woollen stuffs, and as manufacturers, they are on a level much above that of the Greeks around them. Hence, beyond the sphere of Wallachian influence the undecorated style of architecture prevails with a dead and distressing uniformity. The embellishments of wealth- are mostly crude and purely external ; in plan and arrangement the mansion of the Athenian merchant and the cottage of the Aetolian peasant are identical. 1 Cf. Leake, N. G. i. 274. For an account of the Wallachians consult Tozer, Researches in the Highlands of Turkey, ii. 170-182 ; Rodd, The Customs and Lore of Modern Greece, p. 37 fol. CHAPTER V. Kravari ^. The name of Kravari ha:s but an ill sound to modern Greek ears ; it is second only to Agrapha in this respect. Geographically the district might be described as an epitome of Aetolia, in that its mountains are not disposed hap-hazard, but are arranged in parallel lines running from south-west to north-east and alternating with considerable streams, so as to form four distinct zones across the country. And this is the secret of the evils of Kravari. This arrangement of the mountains makes travel and communication very difficult, and the valleys are too narrow to admit of the villages perched on the steep slopes reaping to the full the advan- tages of their inexhaustible water-supply. The soil of the narrow plateaux and of the lower slopes of the hills is good, though scanty ; maize, wheat, and vines flourish. The drawback is that there is not land sufficient to support the population, — (TTevoxapCa, 'narrowness of coast,' is the sum and substance of the complaints of the Kravarites. A large proportion of the population has perforce become pastoral. The mountain pastures of Kravari are therefore not leased to those wandering shepherds who are all included under the vague term Vlachi, whatever their true nationahty, but they are kept in the hands of the villagers themselves. A third resource is found in the forests that clothe the hills. The trees are cut and dressed to standard sizes, and are then transported with infinite toil to the streams. Thus as agriculturists, shepherds, and woodcutters, the ' To Kpd^api, Ta Kpd^^apa. Gell, Itin. p. 295, describes it as ' a high rugged country, but producing good apples ' ! CHARACTER OF THE KRAVARITES. 41 Kravarites win bread from the mountain and valley. Their life is poor and hard, one long struggle with unfavourable natural surroundings, and the effect upon character is plainly marked. The people of Kravari do, not possess that spon- taneous and gracious hospitality which in the rest of Aetolia comes to lighten for the traveller the hardships of the road. Impertinent curiosity is apt to take the place of the eager but respectful interest displayed by the inhabitants of the other Aetolian cantons; and too often a combination of contemptuous indifference and sordid calculation repels the attempt to establish sympathetic intercourse. Of the Pla- taniotes, who are perhaps the most typical examples of the Kravarite character, a saying is current which sums up this attitude, — ' they will pay the stranger to be rid of him.' Platanos is a conspicuous example of another distinguish- ing feature of this district. The poverty of their mountain homes suggests emigration, and it has become a distinct tradition for the younger sons of the Kravarites to seek their fortune abroad, for the most part in Turkey. After years of absence they return, often rich according to a Greek standard, and marrying they spend the remainder of their life among the mountains which they have never forgotten nor ceased to love. If a man, more than usually prosperous, is tempted to remain in exile, his generous gift of land or money for new church or school reminds his native village that his heart is still in Kravari. It is thus not rare to find in Kravari a measure of education and symbols of prosperity such as the physical character of the district would seem to render impossible. Platanos, for example, situated far up the weary slopes of Mount Ardhini, above the narrow vale of the Kakavos, in a desperate wilderness, would seem doomed to remain for ever a miserable village. Yet, in spite of its situation, Platanos is one of the most important towns in the canton^ The houses are large and well constructed of stone, the church spacious, and, after a Greek fashion, handsome, with a London clock which has no companion nearer than Karpenisi or Mesolonghion, the Dan and Beer-sheba of Aetolia. And now the long mule- trains of the Karanghiinidhes are bringing up sand from the > Curiously enough, its commercial relations are chiefly with Nau- paktos. 42 KRAVARI. [CH, Kakavos, to be used in building for the Plataniotes a school that shall be second to none in Kravari'. All these are the gifts of their children far away in Turkey. This patriotism in the expenditure at home of wealth won abroad is indeed common to all the Greeks^, but in Kravari its results are conspicuous. While, again, in the other cantons of Aetolia the traveller hears nothing but Greek, in Kravari the more famihar accents of French, German, or Italian, often greet him as he arrives at some obscure village in the heart of the mountains,— all learnt in that school of modern languages, Constantinople. The Kravarite villages are by no means thinly scattered, but they are generally placed high on the mountain side, so that constantly in order to reach one full in sight a rough descent followed by a fatiguing ascent is unavoidable. Thus many hours of toil often end in but little actual advance. The scenery, however, compensates for the difficulties of the travelhng. The path lies for the most part at a great eleva- tion, leading through forests of oaks or firs. From our eminence the eye follows the windings of the Phidharis or the Mornos imprisoned in the gorge a thousand feet below us. On the opposite ridge the white houses of the villages enliven the sombre green of the slopes. Behind and above all tower the snowy summits of Vardhusi and Kiona, or the distant cone of Chelmos in the Peloponnese. From every point in the land of Kravari the masses of Vardhusi and Kiona form a background to the view, and remind us that we are approaching the eastern limits of. Aetolia. The two mountains constitute one of the most curious features in the physiognomy of the country. Their elevation, over eight thousand feet, gives them an appearance of isolation, and accentuates the contrast between their direc- tion, almost due north and south, and that of the mountains of Kravari. A narrow defile separates Vardhusi from his neighbour on the east, and down this runs the Mega, the ' Great river,' with which the Kokkinos, the ' Red river ' ' This was written in 1893. ^ In 1896 we have a good instance in the magnificent restoration of the Panathenaic Stadion by M. Averof,— a gift which strikingly illustrates what is said on p. 50, note 2. Those who have visited Greece can recall many similar examples of patriotic, if often misdirected, munificence. v] VARDHUSI AND KIONA. 43 flowing along the western foot of Vardhiisi, unites to pro- duce the Mornos. The forms of the two mountains are in striking contrast. Kiona springs skywards with a single effort. Vardhiisi rises gradually in a graceful line from south to north ; we see the snows beginning and gradually deepen- ing and spreading until the eye rests upon the white peak on the northern extremity. CHAPTER VI. Conclusion. The geological structure of the area we have examined is neither very complex nor does it present a great variety of formations, yet in its relation to the external physical features it is a factor of importance in the history of the country. Just as the physical features range themselves in lines parallel from east to west or from north to south, so, geologically considered, is the country made up of three belts of fairly equal width running down from the interior to the coast. On the west, stretching along the Acheloos, there is a belt of sandstone, which extends laterally as far as the base of Kutupas and Kynigu ; southwards it is continued in the Zygos. In the centre, and at each extremity, it is characterized by the inclusion within it of masses of lime- stone, — those of Mount Pteri, the western end of the Zygos above Mesolonghi, and the two excrescences of Varassova and Klokova. South of the central basin the belt no longer keeps to its meridional course, but trends away into the south-east, — a peculiarity shared by the two formations still- to be described. The belt that comes next to the western sandstone is the great limestone axis of Aetolia, the continuation of the Pindos chain. In the north it constitutes the impracticable region of the Aetolian Alps, among which the ridge of Kera- sovon must be regarded as the true prolongation of Pindos. The Une is continued in Chelidhoni : similarly, on the east, KaUakudha follows upon Veluchi. At this point the bi- partite is abandoned for a tripartite arrangement ; for, in a Une with Kaliakudha and Chelidhoni, we have the ridge Missing Page GEOLOGICAL STRUCTURE OF AETOLIA. 45 of Kutupas forming a third member to the west, and en- croaching upon the western sandstone so as greatly to diminish its breadth north of Haghios Vlasis. On the south, Kutupas is connected by Kynigii with Arabokephalon in such a way as to form an unbroken rampart of Umestone across the very centre of Aetolia from west to east, in a direction at right angles to the extension of the forma- tion. The effect of this is seen in the distorted courses of the streams, and the deep gorges which they have cut in the highly tilted strata. From this point this belt also suffers the dislocation noticed in the case of the sandstone. The extension of the limestone is towards the south-east ; at the latitude of Petrochori it is thrown to the east of the Phid- haris, so that it constitutes the greater part of the rugged canton of Kravari. It finally reaches the sea in the pyramid of Rhigani. It is of importance to notice that, in the north, the continuity of this belt is completely broken, namely, between the Kerasovon ridge and Chelidhoni, and between Veluchi and Kaliakiidha. A narrow sandstone band there runs eastwards, and connects the two sandstone regions with each other. The eastern sandstone occupies the space between the central limestone and the two short limestone barriers of Vardhusi and Kiona. It is distinguished from the western sandstone in this respect, that in Mount Oxya it forms a ridge of considerable elevation^, which falls into the series of natural features constituting the eastern frontier of Aetolia. Further south this feature is repeated in the ridge of Trikorphol It is, indeed, largely owing to this that the three regions which we have sketched form a whole, in spite of the sharp severance of one from the other in structure and external aspect. Yet another result has ensued. Intercourse must follow the line of least resistance, and in the case of Aetolia this crossed the Acheloos^ That river, in fact, was a purely artificial line of demarcation between Aetolia and Akarnania, ' 1927 metres. ^ ^552 metres. ' Cf. Vischer, Erinnerungen u. Eindrucke aus Griechenland, p. 520 : ' Akarnanien und Aetolien, welche nicht durch Gebirge, sondern nur durch den Aclieloosfluss von einander getrennt oder vielmehr mit einander verbunden wurden.' 46 CONCLUSION. [cH, and it was characterized by. all the disadvantages of such artificiality. Even before the dawn of history there was the liveliest intercourse between the two districts; that it was throughout of a hostile nature was purely accidental. In a third most striking manner the geological character of the land has influenced its historical development, and in this case also analysis takes us back to the sandstone. The western belt provided an enticing path for Aetolian arms towards the north, partly because it nowhere presents the difficult features of the eastern region, partly because it leads immediately from the central depression. Being thus enabled to turn the barrier of Arabokephalon, the armies of the League were guided by Nature herself into the valley of the Spercheios, and so into the Thessalian plains. It had a deep significance for the history of Aetolia that a gap occurs between the limestone masses of Veliichi and Vard- husi, a gap which is not completely filled even by the sand- stone range of Oxya. Therefore did the League stretch out one hand eastwards to Thermopylai and the Thessalian ports, even as it stretched out another westwards to Ambrakia. Equally operative was the structure of the country in hin- dering the formation of a naval station upon the Corinthian Gulf In the gulf of Krissa, indeed, more than one eligible site was to be fotind, notably that now occupied by Galaxidhi. Yet that port lay too far distant from the centre of Aetolian power, or rather it is precisely with Aetolia proper that communication from Galaxidhi is most difficult ; for, while the pass of Ambliani^ allows easy access to Northern Greece, there is no natural and easy route to be found lead- ing westwards. At the western extremity of the Aetolian coast-line the case is in a curious manner exactly the reverse. There the long hoarded spoils of the Acheloos are deposited, and that maze of shallows and sandbanks created which effectually frustrated the attempt to establish serious com- munication with the outer sea. On the other hand, the connexion of the coast plain with that of the interior is quite satisfactory. Accordingly, contrasting the eastern and western sections of Southern Aetolia, we reach the significant result that on the one hand we have the rich lands around Mount ' See p. 59. VI] NAUPAKTOS. 47 Arakynthos (Mount Zygos) deprived of a sea-board, on the other we have an excellent coast from which the interior is so completely severed as to remain for ever practically non- existent ^. And at Naupaktos, the only remaining eligible site, the latter characteristic is still too predominant. Although the port was perforce adopted^ its value as a factor in the history of the League never reached a really high power ^. It has never, I think, been observed that the value of this town has con- tinuously decreased. The wider the historical relations of Aetolia became, and the more thoroughly economic and non-military their character, the more surely sank Naupaktos into decay. Of greater importance during the seventh and sixth centuries than during the fifth, and during the fifth than during the fourth and the third, Naupaktos has steadily deteriorated. The logical result was reached almost in our own day, when throughout the bitter vicissitudes of the struggle with the Turks its garrison sat wholly inoperative, and its fortress was demonstrated to be a superfluous piece in the game of war ^. If Naupaktos was doomed to a posi- tion of comparative unimportance even in ancient times, how can it be otherwise now when expansion depends entirely upon those economic relations which can never be estab- lished between the town and the interior? The local patriots lament the favouritism, as they love to call it, that prevented the North -Western Hellenic Railway making Naupaktos its point of departure. They resolutely close their eyes to the fact that even the road* which connects them ' Neum. u. Part. Phys. Geogr. p. 163 : ' Kann man die fetten Nieder- ungen rings um den Arakynthos hinter dem unschiffbaren Lagunen- district von Mesolongion ein schones Hinterland ohne Kilste nennen, so lag hier [i. e. in Lokris] eine schOne Kiiste ohne Hinterland.' ^ This does not contradict what is said on p. 336. The Aetolians were bound to make the best of the situation. The acquisition of Naupaktos was all-important to the development of the League, in spite of the draw- backs here sketched. ' A fact which should be taken into consideration in estimating the significance of Byron's work in Greece. It was his ambition to make himself master of Naupaktos. * This road is hardly ever in thorough repair, as, owing to their poverty, the Demes can only spasmodically meet the expense of filling up the gaps caused by the annual collapse of the sandstone. But there is no vehicular traffic between Naupaktos and the lakes, and no demand for 48 CONCLUSION. [ch. with the lake region has yet to prove its claim to be con- sidered a necessary and beneficial construction, such as a poor country was justified in putting early in its pro- gramme of internal development. The effect of the above-mentioned combination of charac- teristics found in the sea-board of Aetolia was somewhat curious. The League was compelled to turn to the Ionian islands in order to supply its deficiency in sea-ports, so that Kephallenia became virtually the southern naval station. Similarly, to the geography was directly due the strange phenomenon that we observe in ancient Aetolia, — there was no capital, no centre of gravity; or, rather, various points external to Aetolia proper served in turn as such centres. Ambrakia, Herakleia, Naupaktos, Lamia, — these were, each in their degree, the real capitals of the League, and as such they were the objectives of its Roman enemies. It was only to a Philip that the coup de theatre of the sack of Thermon could recommend itself as an enterprise worthy of a serious general. It is a mistaken notion that this non-existence of a capital in ancient Aetolia is to be attributed to an impossi- bility of finding a suitable site^, — that there is in Aetolia absolutely no room for such a city. What combination of physical characteristics could be foun,d more favourable than that existing in the very heart of Aetolia? The political influence gathered in the hands of the Trichonian chieftains '^ affords ample proof that some other reason must be sought to explain the absence af great cities in Aetolia proper. That reason we have virtually already given. The towns of Old Aetolia were cut off from all real contact with the sea, and in no less degree from contact with the tribes of Aetolia Epiktetos. Yet intimate connexion with both sea and the interior was an essential condition. If the former desideratum had been frankly abandoned, and if Trichonion or Thermon had been elevated to the position of a capital city, the diffi- culty was not thereby solved; for intercourse between the road, as the agogiats constantly take to the hills to avoid the detours of the highway, not only here but throughout Greece. ^ Cf. Mahaffy, Greek Life and Thought, p. 7. He says that we find no valley or arable land all the way west of Parnassos until we reach Akarnania : that consequently we could not possibly have a capital, as there is nowhere room for a large town. " See p. 235. w o H O 13 o H W < O 2 O O O V'finayFrt:ss. To lajx l^ofi'' 53 THE AETOLIAN TRIBES 'AiravTa fitv aSv ra alviyfiara \veiv in axpi^fs ov paSwv, tov 8e TrXijflous t5)V laiBevoficvav eKTcSevTOs eh to fiiirov, tS>v /iev OfioKoyovvTav ^oKKtjXols 'Toiv o evavTiovfieyaPf einroparepov ttv Tis dvvaiTO elxdCeLV i^ avr&v ToKrjdes. Strabo, p. 474. CHAPTER VII. The Aetolian Tribes. I. We learn from Strabo that Aetolia was generally under- stood to fall into two divisions, known as Old and Acquired (Epiktetos) Aetolia respectively^. The genesis and real meaning of this distinction are obscure, but it must bear reference to the extension of the Aetolian power beyond its original narrow limits, whatever may have been the way in which- that extension was effected. There is, indeed, a real and significant contrast in our earliest notices of the people and their land. Homer knows only five Aetolian cities, — Pleuron, Olenos, Pylene, Chalkis and Kalydon; and their mythic glory is second to none in Greece '\ We search the poems in vain for the names of the OzoUan Lokrians or the Akarnanians, their neighbours on the east and west; of the tribes to the north we should hardly expect to find any mention. Homer's Aetolia is thus nothing more than the plain that extends between the sea and the Zygos, of which the limits on the west and east are the Acheloos and Mount Klokova. That the central plain also was included in Old AetoUa cannot be proved from the Iliad. When we next hear of Aetolia we are in a new world. The Aetolia of Thucydides is the abode of rude savages 3; ^ Str. p. 450 : Kai 817 KOI SijiprjaOai avvi^aive Sixa t^v AhaXlav, Koi T^v fih apxaiav "KiyeaOai Trjv 8' eVwTijTOi'. Cf. id. p. 460. ^ Horn. //. ii. 638 :— AlrcaK&v 8' fiyciTO Boas 'AvSpaL/iOVOs vios, oJ nXevprnv iv€\iovro Koi 'QXtvov rjbe TivK-qvijV XakKiSa T ciyxidKov KaXvSavd re irerprjea-tTav. Cf. id. ix. 529 fol. ' That at any rate was the notion current at the time, but it was far from the truth. 56 THE AETOLIAN TRIBES. " [ch. unwalled villages take the place of the strong and famous cities of the Iliad. Such is New Aetolia, of far wider extent than the Old, and of far different character. Where the one ended and the other began it is hard to say. In default of exact information as to the lines of division between the tribes, geographical considerations mainly must guide us to our conclusions. Exactness is impossible. We shall be fortunate if our utmost endeavours enable us to locate the various peoples in a way that harmonizes the scanty notices we possess. We must not seek to trace their boundaries with a precision that was perhaps unknown to themselves. The grounds upon which we feel justified in extending the term Old Aetolia beyond the coast plain into the central basin are clear from our general survey of the natural fea- tures of the country. It would not be easy to believe that the rich and powerful inhabitants of the maritime plain could have been prevented by the barrier of the Zygos from spreading into the broader expanse to the north. The pass of the Kleisura is a natural gateway inviting communication between the two plains. Nevertheless, it seems equally improbable that the whole of the central basin could have belonged to Old Aetolia. We have shown that there is a sharp contrast between the eastern and western divisions of the depression ^. Strictly speaking, it is only the western part that can be called a plain. The eastern portion is of a different character, the space between Arabokephalon and the Zygos being taken up almost entirely by the lake ; only at intervals, the hills, receding somewhat from the water, leave bays of fertile land, connected one with the other by a rocky path over the intervening spur of the moun- tain. It seems most probable that this system of alternate plain and pass, which guarded, as we know, the approaches to the capital of the League, did not form part of the area included under the name Old Aetolia. The limit of Old Aetolia on this side, if a limit was ever in existence, would fall most naturally some way east of the Kleisura, so as to embrace the western end of Lake Trichonis, with the rich land on its southern shores. If we had only the geography to guide us we should draw the boundary some- ' See pp. 17, 18. vii] NEW AETOLIA. 57 where about the promontory east of Gavalii. Then, skirting the ridge above the modern town of Agrinion, the frontier Hne would run westwards to the Acheloos ^. And this, in fact, is what Strabo tells us of the limits of Ancient Aetolia, as understood by him or his authorities. He says * : — ' Ancient Aetolia embraces the coast from the Acheloos as far as Kaly- don, extending also a good distance into the interior, which consists of a fruitful plain ; in this plain lies Stratos, and also Trichonion, possessing excellent land.' Strabo evidently intends to give us the extreme eastern and western cities of the inland plain. The whole of the remaining country belongs to New Aetolia; but we should be unable to say much about the tribes in possession in historical times had we not the story of the Athenian invasion of 426 B.C. Thucydides lays the foundations of our knowledge in prefacing his account with the meagre description furnished by the Messenians with respect to their dangerous neighbours. 'The Aetolians,' they said, 'though a warlike nation, dwelt in unwalled villages, which were widely scattered; and as they had only light-armed soldiers they would be subdued without difficulty before they could combine. Demosthenes should first attack the Apodotoi, then the Ophioneis, and after them the Eurytanes. The last are the largest tribe of the Aetolians ; they speak a dialect more unintelligible than that of their neighbours, and are said to eat raw flesh ^.' ' We must perforce leave undetermined the relation between the Thestieis, &c., and Old Aetolia. It is possible that the frontier of Aetolia Antiqua stopped short at the southern shore of the lake of Anghelokas- tron (see pp. 87, 173). The Thestieis may have been the intermediaries by v^hose means Aetolia was extended until it ultimately embraced the whole territory inhabited by the three tribes of the Epiktetos. But the part played by the Thestieis in the political development of the country does not justify any inference as to the boundaries of the divisions into which Aetolia was generally understood to fall. " Str. p. 450: apxaiav jxev Trjv airo roi 'AxeX""" M^'XP' KaXi;8£i/of wapaKiav cttI TToXir Koi Tijs /ica-oyaias avfiKovtrav evKapnov re nai irebiabos, g eWl kqi Srparoi xal TO tpiX'i'Viov ap'uTTtjV ^xov yfjv. ' Thuc. iii. 94 : eirixeipeh 8' ixtXevov irparov p.iv 'AiroSaTois, eireira 8e 'O0io- VfCo-i, Kol p.eTa roirovs BipvraiTiv, mrep plyuTTOv fiepor eorl raw AiraXSv, ayvm- a-TOToroi 8e y\S»raav koi ano4>dyoi. ela-iv, as Xeyqj/Tat. The last part of the 58 THE AETOLIAN TRIBES. [ch. Starting from a point to the east of Naupaktos, Demo- sthenes entered Apodotia with the morning hght. Thucydides does not, indeed, expressly say that it was upon Apodotia that the invaders first fell, but it is hardly likely either that he should have failed to tell us, if Demosthenes had disregarded the advice of the Messenians, or that they iri sketching the plan of campaign should have abandoned the true geo- graphical sequence of the territories to be attacked ^ The Apodotoi, therefore, occupied the country to the north-east of Naupaktos ; but they had no sea-board. This is what we infer from the words of Skylax, who says that ' Aetolia stretches along the whole length of the interior of Lokris, and as far north as the Ainianes ^' And Dionysios, son of Kalliphon, says: — 'the Lokrians dwell south of Aetolia ^' The inference is confirmed when we examine the march of Eurylochos the Spartan upon Naupaktos, immediately after the repulse of the Athenians. Before leaving Delphi, Eury- lochos is very careful to secure hostages, by force or persuasion, from all the Lokrian towns, from Amphissa to Oineon and Eupalion, for ' through the Ozolian Lokrians lay the route to Naupaktos*.' This conchisively proves that the coast from Amphissa to Naupaktos belonged to the Lokrians. Their next neighbours, at no great distance from the sea, at least in the neighbourhood of Naupaktos, were the Apodotoi. Between the coast-line and the course of the Mornos there is interposed the range of Trikorpho and Vigla ; what more satisfactory than to see in it the boundary between the Ozolian Lokrians and the Apodotoi ^ ? At any rate, if the frontier was a natural one, none other is possible. How far the Apodotoi extended to the east or the west we have no means of knowing exactly. The idea of Demosthenes was to ' proceed through the Ozolian Lokrians to Kytinion description looks like an item derived from some other source than the Messenians. ^ Or some may prefer to draw our conclusion from the implication in Thuc. iii. 96 : tSWo KaTaiTTpe^afUvos ovras eVi '0(j)i.oveas, k.t.X. '^ Skyl. Cary. § 35 : 'H 8« AiTmXio irapriKei rifv AoKpiSa iraa'av djrA /leo-oyeias liexpi AlvidvcDV. ' Dion. /. 70 : OStoi KaroiKoviriv 8e Trpor pftrrjii^piav IdratXlas. ' Thuc. lil. loi : iTreKtipvKeiero EvpvKo\os AoKpois rois 'OfdXaW 8ia rovrav yap ij oSos ^v is NavwaKTOv. ° Bazin suggests this, M^m. p. 303. vii] OPHIONEIA. 59 in Doris, keeping Mount Parnassos on the right, until he came down into Phokis ^.' It is clear that he intended to take the road that runs northwards from Amphissa through the pass Ambliani ^. This pass divides the spurs of Parnassos on the east from those of Kiona on the west, and descending directly into the valley of the Kephisos comes to an end at the famous khan of Gravia. The eastern boundary of Apodotia fell, therefore, somewhere to the west ,of this pass. Mount Kiona itself would undoubtedly form part of the line ; but we cannot say to what extent the Apodotoi encroached upon the valley in which are the modern towns of Lidhoriki and Malandhrino. On the west, the sudden southward bend of the Mornos between Mount Vigla and the range of Makryvoros would furnish a natural limit to their territory, if we could posi- tively affirm that they were confined entirely to the left bank of the river. We have, however, no evidence as to the extent of Apodotia to either the west or the north, and it is quite impossible to prove that the Mornos was actually the boundary on those two sides. All that is certain is that, to the north, the next neighbours of the Apodotoi were the Ophioneis. This is implied in the somewhat obscure sentence in which Thucydides explains the plan of campaign in the mind of Demosthenes. He had halted at Teichion, — ^which, until the contrary is proved, we must suppose to be in Apodotia, — and from Teichion he sent back the spoils to Eupalion in Lokris, ' for he did not intend to attack the Ophioneis yet: when he had subjugated the rest of the country he would make an expedition against them subsequently, after his return to Naupaktos, if they continued to resist ^' This sentence enables us to make out the exact point to which Demosthenes had penetrated.. ^ Thuc. iii. 95. ^ 'A/iTrXtavij. This pass was naturally of great importance during the War of Independence. See Trik. iii. 151. We still see in it the ' tam- biiria' raised by Panuri^s. For the story of the khan of Tpa^ia, see Trik. i. 265 fol. ' Thuc. iii. 96 : koI alpel . . '. Teixiov, efieve re aliTov koX rrjv Xciav is Evna\iov TTJs AoKpiSos airewefi^fV rrjv yap yvajxufv eixe riWa KaTaa-Tpeiffdp^vos ovtos iirl '0iov KoKovfievov Spovs' Kal peiv avTodev els dahairarav Stiritfp ttvov, Te6d(j>6ai 6' if Tffl opa. Tourm Neatrov Tou KevTavpov, ov 'VIpaKKtjs oiriKreivev. In margine haec, tntpteiov oTi to irjfyavov rfv n (jjapiiaKOf. These are the springs called Bpaiiovepd. See Pouq. Voy. iv. 8. They do not now at all answer to their ancient repute. ' Plut. Quaest. Gr. xv. See p. 329. * Str. p. 427 : hiTtokiKov itoKlxviov. ' Ibid. : EoTi 8e vvv PdraiKav iCKiinrov irpoaKpivavros. VII] COAST BETWEEN NAUPAKTOS AND CHALKIS. 65 So also Pliny ^ and Pomponius Mela ^ are speaking, whether they know it or not, of the period of the supremacy of the League in reckoning Naupaktos to Aetolia ; they are follow- ing Skylax, who of course gives the state of things as it was after 338 b.c.^ (3) Lastly, in his thrice repeated statement that Antirrhion formed the boundary between Lokris and Aetolia, Strabo is referring to the arrangement existing in Roman times *- The same is true of Ptolemy, when he assigns Molykreion to Lokris ^. It is clear, then, that before the extension of the League obliterated all boundaries the Aetolians did not possess a foot of ground east of Mount Klokova. This coast was all originally Lokrian, the Corinthians and Athenians being temporarily in possession of its western section, together with the territory belonging to the neighbouring Aetolian city of Chalkis, during the greater part of the fifth century before our era. The Ophioneis, therefore, were confined to the interior, north of the strip of Lokro- Athenian territory which stretched along the Gulf between the rivers Phidharis and Mornos. It is, of course, impossible to trace the exact line of demarcation, but the geographical features are sufficiently strongly marked to give us a fairly probable frontier. We see how closely the Mornos and the Phidharis approach each other at the angle formed by each of them in the second change of direction, — from south to south-west in the case of the Phidharis, from south-west to south in the case of the Mornos. We imagine that the Ophioneis occupied the whole of the intervening district, behind Mount Rhigani. On the west, the Phidharis itself must have constituted the frontier of Ophioneia. It is possible that the order of narra- ' Pljny, H. N. iv. 3 : Sed in Corinthiaco sinu oppida Aetoliae Naupac- tum Pylene. ^ Pomp. Mela, De Choro. ii. 43 : in Aetolia Naupactos. ' Skyl. § 35- * Str. p. 336 : tA 8' 'Aurippwv iv iieBopiois rrjs AlroKlas Kal Trjs AoKpiSos. Ibid.: etff e^ijs em to 'Avrlppiov AItoXoL Id. p. 460 : tA 'Avnppiov to rrjs AkaXias Spiov Kal T^f AoKpiSos, ^ Ptol. Geogr. iii. 14. 3. He also gives Antirrhion ('Avrippiov uKpov), as well as MoXvKpia, to Lokris. F 66 THE AETOLIAN TRJBES. [ch. tion observed by Strabo ^ points to this conclusion as to the limits of the tribe. After saying that the Euenos (Phidharis) rises among the Bomieis (a section of the Ophioneis), he passes at once to the maritime plain ; from which we might infer that no tribe other than that of the Ophioneis, with its various subdivisions, interposed between the two points ^. A more valuable argument is the conjecture of Leake, that the name of the Ophioneis still survives in the word Phidharis, which is apparently derived from t6t, the Romaic form of o(^is (snake) '. There is no positive evidence from which to derive the line followed by the northern frontier of Ophioneia. As the Euenos rose among the Bomieis, we shall probably not be far wrong if we imagine all the country round the head waters of the Phidharis within the angle between the Oxya range and Mount Vardhiisi to have belonged to the Ophioneis ; but there is nothing in our authorities to indicate how far they extended to the west, or whether in this direction they were confined to the left bank of the river. It is tempting to assume that the gorge of the Phidharis cutting across the country from east to west formed the northern limit of their territory; equally tempting, and perhaps more probable, is a frontier formed by the eastern continuation of Arabo- kephalon towards the Oxya mountains. This would agree with the connexion which we have pointed out as existing to-day between the villages on that watershed and those of Kravari *. Wherever exactly the frontiers of the Ophioneis fell, their land was, in the main, co-extensive with the cantort of Kravari, which is severed so curiously from the other cantons of modern Aetolia. ^ Str. p. 451 : d ¥>Tjrivos irorafibs ap)(eTai fi€v eK Baifiteoiv rav iv ^Ocj^ievatv AlraiKiKa idvei . . . pel 8' ov 8ia rrjs Kovpriniirjs, K.r.X. " I let the argument stand, as I find that it has been used by Becker, Diss. iii. 15 : but it assumes for Strabo a greater measure of completeness than I can credit. ° Leake, N. G. ii. 625. The Ophioneis are, therefore, the ' Snake-men ' ; certainly an indication of totemism. Cf the inscription from Kryonerii, p. 199. Ophis is known as the name of a river. See Paus. viii. 8. 4. Thucydides uses the form Ophioneis, but the name of the tribe appears in the shorter form Ophieis in Strabo, pp. 451, 465. And, in fact, from the inscriptions quoted on p. 76, note 8, the form ■0(/)i6ur would seem to have been that usually employed, at any rate during the second century B. c. * See p. 23. vn] THE KALLIEIS. 67 It is well known that we can speak about the Ophioneis in greater detail than about any other Aetolian tribe. Thucy- dides has recorded the names of two of their minor cantons, —those of the Bomieis and the Kallieis, — though without giving us any information as to their relative position^. From other sources, however, we can approximately fix' their abode. Having failed to force Thermopylai in 279 b.c, Brennos detached a strong force to effect a diversion in Aetolia, hoping thereby to cause the Aetolian contingent to withdraw from the defence of the pass. Orestorios and Komboutis led forty thousand men back across the Spercheios, and then westwards along the valley, until they dashed suddenly into Aetolia against the capital of the Kallieis ^. The object of the Galatai was to fall upon the Aetolian land as soon and as effectively as possible. We may therefore conclude that the territory of the unfortunate Kallieis lay nearest, both to the pass of Thermopylai and to the Thessalian border. The requirements of the case are met exactly if we place the tribe upon Mount Oita and the spurs of Kiona overlooking Doris. Whether it was the Kallieis or the Apodotoi that occupied the valley of the Mega must be left to conjecture. My own view is that the two tribes came into contact in the neighbour- hood of the modern town of Lidhoriki, the Apodotoi being confined to the left bank of the river ■'. Adopting the ordinary phraseology*, we have spoken of Mount Oita as the home of the Kallieis, but this does n,ot seem strictly accurate. The Kallieis appear to have given the name Kallidromos to that part of the range which belonged to them. A certain amount of confusion seems to have existed even in ancient times in connexion with Kalli- dromos. We have two statements about it, in the pages of Livy and Strabo. The .former describes the range of Oita ' Thuc. iii. 96 : 01 ea-xaroi 'O(j)ioveav . . . Baiurjs Koi KaWi^s. ^ Paus. X. 22. 3 : ot oir'ura Kara tov 'S.nepXfiov ras yecjiipas Koi aSBis 8ia Oetra-a- \ias oSevtravres eiiPaWovcri.v is Trjv AtrtoXiav. Kal to h KaXXitas Kofi^ovTis oi ipyairajievoi xai 'OpecrrSpLos ^aav. ' This view is necessarily connected with my theory of the historical relations of the components of the Aetolian name, and cannot be further discussed here. * E.g. Bursian, Geogr. i. 142. F 2 68 THE AETOLIAN TRIBES. [ch. thus ^ : — ' These mountains divide Greece in the middle, just as Italy is divided by the ridge of the Apennines . . . They extend from Leukas and the western sea through Aetolia to the other sea on the east, and are so rough and precipitous that not only troops, but even ordinary travellers, find diffi- culty in crossing them by the few paths they afford. At the eastern extremity they are called Oita, and their highest summit Kallidromon, at the foot of which, towards the Maliac gulf, there is a pass . . . called Pylai, and sometimes Thermo- pylai.' Strabo has the same account ^. ' Oita,' he says, ' extends from Thermopylai and the east as far as the Ambra- kian gulf in the west. We may describe it as cutting at right angles the range that stretches from Parnassos to Pindos and the northern tribes. That part of it which inclines towards Thermopylai is called Oita : it is two hun- dred stades in length, craggy and lofty, reaching its greatest elevation at Thermopylai, where it runs up in a peak, and ends in sheer precipices towards the sea^. A narrow passage is left, through which one may enter Lokris from Thessaly by the coast route. This passage goes by the names of Pylai, the Pass, and Thermopylai, — the last name being derived from the neighbouring hot springs, which are sacred to Herakles. The mountain above the pass is Kallidromon. Sometimes this name is applied to the remainder, that is to say to that part which runs through Aetolia and Akarnania to the Ambrakian gulf*.' It is clear from the above passages that Kallidromos, as understood in the more restricted sense, corresponds to the modern Saromata, the ridge that extends above Thermo- pylai between Spartia (Mount Knemis) and Katavothra (Mount Oita). Both Livy and Strabo are in error in regard- ing the summit above the pass as the highest point of the ' Livy, xxxvi. 15 : ab Leucate . . . per Aetoliam . . . tendens. Extremes ad orientem montes Oetam vocant, quorum quod altissimum est Calli- dromon appellatur. " Str. p. 428. The origin of the two descriptions is clear. ' TovTov hi] TO liiv wp6s Gfp/iOTTuXas veyevxtis fiipos Oirij KoKebrai arahiatv Sio/co- iriaiy to firJKos, Tpaxv Ka\ injfri\6v, vyjrriXoTaTov 8c Kara ras Ofp/jurnvKas' Kopu0ovrat yap evTavda koX TeXfvm Trpos o^els Koi dn-OTd/ioDf nexpi Trjs BdXaxTirqs Kpr]p.vois. * TO S' virepKeifievov opos KciWiSpopov ' nvis Se Km tA XoiTrAv tA &i AlTcaXias Koi Trjs AKapvavtas Sirjicov peXP' '"''" 'Ap^paxiKov koXttov KaXXiSpo^oK irpoir- ayopeuQVO't, vii] MOUNT OITA. 69 range. The stupendous precipices called Katavothra, above Hypati, are much higher than those overhanging Thermo- pylai; and Mount Patriotiko, the loftiest summit in the Katavothra range, is nearly a thousand metres higher than Saromata V We see from Strabo that the name Oita was strictly limited in its proper application; the two hundred stades assigned to it is very nearly the distance in a straight line between the pass of Thermopjdai and the valley of the Vistritsa. It is easy to understand how the hills above the pass should have come to bear a separate name. They are severed from the chain of the Katavothra mountains by the gorge ^ of the Asopos (the modern Karvunaria), so that their most intimate connexion is not with that range but with the mountain country at the head of the Kephisos valley. Oita seems, therefore, to have been in reality not the general name of the whole range running eastwards from Pindos to the Maliac gulf, but that applied to the striking precipices overhanging the plain of the Spercheios between the rivers Vistritsa and Karvunaria. The general name of the range, lying behind those precipices, was Kallidromos ^. Naturally, both names were often used in a manner not strictly accurate. Oita was extended so as to include the whole mountain region lying between the Spercheios and Kephisos valleys eastwards as far as the Epiknemidian Lokrians. Similarly, the name Kallidromos was, on the one hand, unduly extended to the whole central line of mountains between Leukas and Thermopylai ; and, on the other hand, unduly restricted to the summits immediately above the pass itself*. Between the name of the mountain Kallidromos and ' Cf. Leake, N. G. ii. 9 fol. ^ Cf. Herod, vii. 199 : Siaa^a^ npos tuaafi^pirfv Tprixlvos, 8ia 8e Trjs Siaatjilayos 'Aa-anros woTajws pen. ' Ptolemy is therefore strictly correct in saying that the Euenos rises in Kallidromos. Dionysios, son of Kalliphon, who says that it rises in Pindos, is also not far wrong, though it is scarcely justifiable to extend the name Pindos so widely. See p. 93. * Cf. Leake, N. G. ii. 63 fol. It is in this third and narrowest sense that the name is used by Livy (xxxvi. 16) : duo milia Heracleae substiterunt ; duo vtrifariam divisa Callidromum et Rhoduntiam et Tjchiunta— haec nomina cacuminibus sunt — occupavere. We learn from what immedi- ately follows that these heights were fortified (castella Aetolorum). See also App. Syr. 17 fol. 70 THE AETOLIAN TRIBES. [ch. that of the tribe of the Kallieis there is evidently a connexion, which perhaps survives in the modern name of the great mountain Kaliakudha, the central point of the ancient Kalli- dromos as used in its widest sense. With respect to the boundaries of the Kallieis towards the east, south, and north, we have already said all that is pos- sible ^. On the west of their territory a somewhat greater precision is attainable. The long line of Mount Vardhiisi must have separated them effectually from the rest of the Ophioneis, including the subdivision of the Bomieis. At the southern end of Vardhiisi, however, the Kallieis, if we cor- rectly assume them to have occupied the Mega valley, would come into contact with the main tribe of the Ophioneis ^ We have already seen that the Bomieis inhabited the country round the sources of the Phidharis ^. It agrees best with the configuration of the district to imagine that the tribe extended southwards to the crest of the watershed between the head waters of the Phidharis and the Kokkinos. On this ridge, which runs south-west from Vardhiisi to the Vlacho- viinia, the Bomieis would be conterminous with the main tribe of the Ophioneis. They would naturally spread west- wards along the valley of the Phidharis, between Mounts Trekuri and Ardhini on the south and the continuation of the Arabokephalon on the north. From the Oxya mountains, therefore, to the confluence of the Vasiliko with the Phi- dharis, the Bomieis were conterminous with the Eurytanes. We need not conclude from the words of Stephanus * that the country of the Bomieis was exceptionally mountainous ; nor indeed would such conclusion be in accordance with fact, ' See pp. 62, 67. ' For the sake of completeness only I notice here the old error about the Aetolian Tripolis. Stephanus writes : KaXXmi, m\is /lia rjjs iv AiVwXici TpnrdXea)s. We must correct A?T<»\ia to 'ApxaSia. The copyist had in his mind Paus. viii. 27. 4 : npo(reyiveTO &e Koi TpijroXis ovofia^oiiivt], KaXXi'a Km Ahroiva koi HavaKpis. The Statement in Stephanus cannot be saved by enumerating the three tribes of Aetolia as members of the Tripolis, for Aetolia was a union of Tribes, not of Cities, to which latter union alone the term Tripolis is rightly applied. Nor, again, can we hold the Tri- polis to have been constituted by the Ophieis, Bomieis, and Kallieis, as it is evident that the last two were merely sections of the Ophieis. ' See p. 66. * Bcofioi, \6(j)oi AlroKias " 01 KaTOiKOvvres Bra/xicir, vn] THE BOMIEIS. 71 unless they covered a much greater area than we have assigned to them. It is also impossible to bring the name of the tribe into any connexion with the myth of Herakles^. The legendary scene of that hero's death was near Herakleia, perhaps outside the limits of Aetolia proper ; at best it would be among the Kallieis. The only probable explanation is to regard the name as originating in some fancied resemblance between the hills and altars. It is true that in the land of the Bomieis we find semi-isolated conical heights which might have suggested the comparison ^. Nothing, however, is more difficult than the attempt to recover for a moment that power of perceiving likeness between widely different objects which seems to be a prerogative of youth. The imagination of later generations is less naive and lively, so that resemblances once striking and obvious can no longer be perceived. It requires a special effort in order to understand and delight in the imagery of poetry belonging to a race not our own, and distinctions of nationality find expression to no slight extent in the choice of epithets. Instances in Greece of names that had their origin in this delicate feeling for resem- blances which for ever escape our perception are numerous. The Akontion ridge above Orchomenos in Boiotia, the Harma in Parnes, Mount Othrys in Thessaly, the Kynos- kephalai hills upon which Philip of Macedon suffered defeat, are only a few examples. The explanation in Stephanus, therefore, adds nothing to our knowledge of the country, though it may preserve to us an Aetolian term, and possibly a glimpse into Aetolian mythology. ' Cf. Kruse, Hellas, ii. 229. '' This is owing to the fact that the tract inhabited by the tribe is the southern extension of the eastern sandstone belt, between Vardhusi and the limestone hills of central Kravari. Grasberger, in his Studien zu den Criechischen Ortsnamen, pp. 116, 155, renders 01 Bwjuot, 'der Stufenberg.' CHAPTER VIII. The Aetolian Tribes. II. It is evident that the Messenians represented the Eury- tanes as inhabiting the country beyond the Apodotoi and the Ophioneis. In his sketch of the proposed system of Athenian operations, — first the subjugation of the Apodotoi, and then the campaign against the Ophioneis, — we notice that Thucy- dides does not mention the third tribe. It is probable that the omission is not without design. It may be an intimation that Demosthenes had begun to realize the magnitude of his task, and to realize also that the conquest of Eurytania was not essential to the completion of his scheme of invading Boiotia. It would be strange if so capable a man could have remained long in the dark as to the real necessities of the case. His design, was, for that time, a bold one, possibly rash and unfeasible, but it was essential to its success that he should secure the passes into Doris and Phokis by the conquest of the Apodotoi and the Ophioneis, who occupied the mountains above them in the west and north. From the Eurytanes there was nothing to fear, so far as their geographical position was concerned ; they were in nowise contiguous to the route of the expedition. The real danger, the rock upon which the Athenian general made shipwreck, was the fact that the political organization of these sup- posed savages supplemented the natural advantages of their mountain strongholds in a way that rendered them invincible. The results reached in the preceding Chapter make it clear that the Eurytanes must be placed to the north and north- TRIBES ON THE NORTH OF AETOLIA. 73 west of the Ophioneis. The OxyA range, running north-west from Vardhusi to Veluchi, formed their frontier on this side ; for the Ainianes lay upon its eastern slopes and in the valley of the Spercheios. So much can be said with certainty; but in finding the limits of the Eurytanes on the north, west, and south, we are left to conjecture. The area that they inhabited must have been considerable, for there is no doubt that their name bore direct reference to their terri- torial dispersion^. We may, however, perhaps suspect its genuineness as an Aetolian ethnical term I We have a long list of peoples more or less closely con- nected with Aetolia in the north. Pliny, for example, mentions the Athamanes, Tymphaei, Ephyri, Aenienses, Perrhaebi, Dolopes, Maraces, and Atraces. He calls all these 'Aetolian V but that could only be justified, if at all, by supposing him to refer to the time at which the power of the League was at its height. Topographically, at least four of the tribes in his list have no connexion with Aetolia. From Strabo we get the tribes on the west, — the Amphi- lochoi, Aperantoi, and Agraioi, — with respect to all of which there is the difficulty of distinguishing whether they are real divisions of the Aetolian name, or only tribes added by con- quest. Of most of these tribes on the north and west little more than the name has survived, and it is hopeless to attempt to fix their habitation in any but the vaguest manner. They appear to have been offshoots from the people of Thessaly or of Epiros ; the Eurytanes themselves seem not to have been without a strain of Epirot blood. A careful examination of our authorities soon reveals the fact that the tribes on the north of Aetolia reduce them- selves to four *, which are generally mentioned in groups of ^ In spite of what we read in Tzetzes on Lyk. Kass. I. 799 : 'Apto-TorcXijs (j)tjv nvas. Strabo, p. 427 : toIs 8' 'ETrtKvij^iifiiow Alviaves o-u>iE;feis 01 rrjv OiTrjv exovres. See also Leake, N. G. ii. 21. * Pol. X, 42 : 6 Alviav koXttos. ° Str. p. 430 : (X^i fi' Tj fiev *6i£tis to yoxia to napa ttjv Oirriv airo tov MoKiokov KoXirov Koi TivXa'tKOv p-exp^ TJjy AoXoTrias Kal rrjs UlvSov duvreipovra. Id. p. 433 : 6p6pov 8e ra Tv(\)pij(TT(S Koi tois AoXo^iv. It is true that in this same section Strabo speaks of Typhrestos as belonging to Dryopia: jas irrjyas exovros (sc. TOV ^Trepxetov) sk Tti0p))(rTOu ApvomKov opovs. But ApvoniKov must be changed to AoXomKov : Cf. Bursian, Geogr. i. 87, note 5. For the con- nexion between Dolopia and Pindos, see also Strabo, pp. 432, 434, 437, 440, 450. AoXoTTi'a (AdXoTres) K.ai r\ Ilii/Sor is the, regular expression. ° Horn. //. ix. 484 : NnTop 8' ecrxan^v iSirjs, AoX6miroL, however, never rose to any importance, and seem in fact soon to have been absorbed by the warlike tribes surrounding them : at least they are never mentioned in any of the military operations of which this part of the country was the theatre. Still, it is probable that Strabo, in speaking of the Perrhaiboi as neighbours of the Aetolians, was thinking of this offshoot from the main body. The Athamanes ga:ined some measure of importance in the history of the later years of Greek independence *. On the east, in the chain of Pindos, they came into contact with the Thessalians ; on the west they extended into Epiros *. This is proved by Livy's account of the Roman advance from the coast into Thessaly against Perseus «- They ' Str. p. 450 : VTrepKeivTai 8' iv rfj /ifo-oyaia Koi rots irpotr^opeiois fiepetri rav piev 'AKapvdvav ' A/j^iXoxoi, tovtcov hi AoKoires kcu. fj Hlvbos, t&v 8 AlraXav Tleppai^oL T€ Kol 'ABapaves KaX Amaviav n pipos tS>v Trpi O'ttjv exovrmv. ^ Str. pp. 440-442. Cf. Homer, //. ii. 749 fol. See Leake, N. G. iv. 3"- ' Str. p. 440 : 01 pev ovv Ufppca^ol . . . «s Trjv opeivrjv caraveaTrjcrav 01 TrXeiou? rfjv nepX Ilivdov koi 'ABapavas xal AoKonas, Id. p. 442 : to 8e rroKv pipos els ra rrepl rfjv 'Adapaviav opr) xai Trjv IllvSop i^in«re- vvvl 8e piKpbv fj ov8ev avTav tx""^ o-mfeTai. Cf. Leake, N. G. iv. 213. ■• Str. p. 427 : 'Adapaves 8' vototoi t&v 'HireipaiTSiv els a^'uopa npoaxBevres, rj8if) T&v akXav dTreipriKOTcav, Kai per 'ApvvdvSpov tov ^a(riKea>s Svvapiu KaToa-Keva&d- pevoi. = Cf. Leake, N. G. iv. 212. ' Livy, xlii. 55. 76 THE AETOLIAN TRIBES. [ch, marched first through Epiros, and then through Athamania to Gomphoi, which fortress commanded the entrance into Athamania from the side of Thessaly ^ On the south their territory extended to the frontiers of Amphilochia and Dolo- pia^. Livy tells us that Philip gave Zakynthos to King Amynander as the price of a passage with his army through Athamania into Upper Aetolia *- From Ptolemy we get the idea that the Athamanes lay somewhat to the north-east of the Amphilochoi *. The tribes on the west of Aetolia form a triad, constituted by the Amphilochoi, Aperantoi, and Agraioi. The Amphilochoi, on the west and south-west, came down to the shores of the Ambrakian gulf. On the north Athamania and Epiros, on the north-east Dolopia, were conterminous with them. Akarnania fell to the south, and the land of the Agraioi to the east and south-east. The difficulty with regard to this tribe is not one of geography. Polybios seems to reckon them as true Aetolians. In the conference of Philip with Flamininus and the Roman allies at Nikaia, in 198 e.g., Philip exclaims in anger to the Aetolians : ' Why, most of the Aetolians themselves are not Greeks; for neither the Agraioi nor the Apodotoi nor the Amphilochoi are counted as Greek ^.' Polybios himself speaks of the Ambrakian gulf as ' extending a long way into the interior of Aetolia^,' which would seem to imply that Amphilochia was part of Aetolia ''. It is impossible to accept the Amphilochoi as an Aetolian tribe on such slender evidence. The mention of the Apodotoi sounds strangely after the long silence since the days of Thucydides. It serves to prove that the old cantons were still kept up in some way alongside the League*. If the ^ Livy, xxxi. 41 : Gomphps . . . imminet Athamania huic urbi. Leake, N. G. iv. 522. ^ Str. p. 450. ° Livy, xxxvi. 31 : earn mercedem Amynandro dederat, ut per Atha- maniam ducere exercitum in superiorem partem Aetoliae liceret. * Ptol. iii. 13. 7 : ' ti.ii<^CK6\a>v, i>v fltrlv 'Ada/iaves avaroKikwTfpoi. ' Pol. xviii. 5 : t6 yap rav 'Aypaav eBvos Kai to rav 'ATroSoTSw . . . oi/c ttrriv 'EXXar. * Id. V. 5 • *'f '■"I'f pf(TOyaiovs av^Kfi tottovs ttis AlrcaiKias. '' Or the words riys AlraXias are used loosely, because some part of Akarnania was subject to the League. * Cf. Sammlung-Collits, No. 1862 = W.-F. 197 : ov inpiaro rrapa UoXepApxov viii] THE AMPHILOCHOI NOT TRUE AETOLIANS. 77 Apodotoi are to be regarded as non-Greek, the Ophioneis, and certainly the Eurytanes, must go under the same category ; and thus we get back to the old worthless tradition, which was jealous of allowing the backward peoples of Western Greece to call themselves Hellenes^. The truth is that Philip, or Polybios for him, used the first names that came to his tongue, — hence the alliteration. An expression used in the only other place in which Polybios mentions the Amphilochoi does, in fact, directly contradict the supposed implication of the passage above quoted. In 191 b. c. Philip, nominally in the interests of Rome, overran the districts of Athamania, Dolopia, and Aperantia ^ ; but in 189 b. c, when the attention of Rome was diverted to the campaign against Antiochos in Asia, the Aetolians seized the opportunity to restore Amynander to his kingdom, and to ' annex once more ' Aperantia and Amphilochia, in order to secure themselves against attack from the west^. In Amphilochia popular feeling was with the Aetolians, as was also the case in Aperantia. They next marched into Dolopia, where 'a show of resistance and of keeping loyal to Philip * ' was made, but soon the Dolopes also threw in their adherence to the League. Subsequently, we find Philip intriguing in Rome against the Aetolians, 'looking upon himself as wronged by their having taken Athamania and Dolopia from him ^.' It is significant that Philip confines his pretensions to Athamania and Dolopia. It was, in fact, possible to make out some sort of case in defence of his claim to those two Tov Teuravhpov '0Uos, date 176/5 B.C.; and No. 1978 =W-F. 313, where the principal is 'AycorpoTos'O^teur. This is dated by the second Strategia of Thoas=i93 b. c. Both these inscriptions come from Delphi. ^ Cf. Str. p. 449. ' Livy, xxxvi. 33 : Inde Dolopiam et Aperantiam et Perrhaebiae quas- dam civitates recipit. Id. 34 : Philippum autem . . . non solum urbes sed tot jam gentes, Athamaniam Perrhaebiam Aperantiam Dolopiam, sibi adjunxisse. See also id. xxxix. 28, 34. Plut. Flam. xv. : 'Ayo/ievav 8e km (j^fpo/iAvfOV wro TOV MaKtDovos tovto juev Ao\6wtov Koi MayvijTiov, toUto Se ABa/id- V(ov Kai 'AjrcpaiTav. ' Livy, xxxviii. 3. Pol. xxi. 25 : ol AItoKoI vop'uravTes ex"" ei'^>'5 Kaipov npos TO TTjv ^ A-fiCfiiKoxi'iv Koi ttjv 'AirfpavTiav avaKTrjaacrdai k.t.\. * Ibid. : TTiprj(TavTes rfiv irpos 0lXm7rov irianv. ^ Pol. xxi. 31 : 'Ekeii/os yap hoKmv aSUats vtto rav AiTrnXSi; a 'Aypaia* Xeyoi/rat be Kai 'Aypaels, as 'Eparocrdtvrjs . So Thuc, iii. Io6, has ttjs 'Aypqiav ; but in chap. Ill, is Trjv 'AypatSa, ° Pol. xviii. 5. See p. 76. * Str. p. 451 : €V 'Oi^ieCo-iy AhaXixa Wvei, Kad&nep Koi o\ ^vpvravis mX 'Aypaioi Koi YLovprjT£S Kol SKKoi. Id. pp. 449, 465* G 82 THE AETOLIAN TRIBES. [ch. at the period in which the Aetolians had already developed a working Federation the Agraioi were still under kingly- rule '. It is quite futile to attempt to argue away this plainly asserted fact, or on the other hand to use it for the purpose of throwing light upon the early history of Aetolia. Thucydides gives us valuable information as to the locality occupied by the Agraioi. In 426 b. c. Eurylochos and the Peloponnesians hastened across the Acheloos with the object of effecting a junction with the Ambrakiots. From the territory of Limnaia, now Karvassaras, at the south- eastern angle of the Ambrakian gulf, the army struck to the right : ' at last they left Akarnania and reached the friendly country of the Agraioi^.' From the Agraia they crossed Mount Thyamos into the plain of the Amphilochian Argos. After their defeat at Olpai, 'the Peloponnesians escaped into the neighbouring country of Agraia, and were received by King Salynthios, who was their friend ■''.' This makes it quite clear that the Agraia lay upon the right bank of the Acheloos. From the position to be assigned to the neigh- bouring states of Akarnania and Amphilochia, the Agraioi can only have occupied the large angle which the Acheloos makes between Sivista and Preventza *- Thucydides, again, is equally clear about the course of the Acheloos, which ' rising in Mount Pindos, and passing first through the territory of the Dolopes, Agraioi, and Amphilochoi, and then through the Akarnanian plain,' flows by Stratos, and so to Oiniadai ^ This seems to be the source of what Strabo says, that 'it flows from the north and Pindos to the south, through the Agraioi, an Aetolian tribe, and the Amphi- lochoi ^' The contention that the Agraeis extended also to the left bank of the Acheloos must be supported by stronger ^ Thuc. iii. iii. Thuc. iii. 106 : icai iwe^r/a'av Trjs 'Aypatav, ouKCTt 'AKapvavtas, i\ias 8e vyov is Trjv 'AypatSa, Sp,opov oSBev fiev napa ^rpdrov ttSXiv, es BdXaa-a-av 8' e'liEir irap' OlvidSas, k.t.X. Str. p. 449 : pecrov exovrfs tov 'ApfeXaov iroraphv peovra djrA tSiv Spierav (tai T^r TItvSov irpos v6tov Sta te 'Aypaiav AlraXiKoii cflyouf Km ' Ap(j}i\Ax(OV. viii] THE AGRAIOI AND THE APERANTOI. 83 evidence than tKat tortured from these words. The Ache- loos does actually flow, as Thucydides says, through Dolopia, Agraia, and Amphilochia; but it does not bisect them, as some interpreters would compel us to believe. It is, in fact, impossible to find room for that section of the Agraeis which is supposed to have dwelt upon the east of the river, except upon a clumsy and gratuitous hypothesis. South of Aperantia, the only territory on the east of the Acheloos into which the Agraeis might have come is the tract intervening between the river Zervas and the central plain of Aetolia, which last no one imagines to have belonged to them^. Now this strip was certainly in the possession of the Thestieis, who cannot be proved to have been a section of the Agraeis nor yet of the more widely extended Eury- tanes. Most probably, as we shall see later, they belonged to neither people. There only remains for the supposed eastern Agraeis a position between the Aperantoi and the Dolopes, one which has as little in its favour as the former. The structure of the country makes it incredible that the dislocation that would ensue upon this theory ever existed. For clearly, unless the territory of the Aperantoi is to be reduced almost to the vanishing point, we can hardly locate the Agraeis elsewhere than in the district that we have supposed to belong to the Dolopes, and the Dolopes them- selves must then be pushed farther to the north and towards the.confines of Thessaly. The Agraioi would thus lie in a line stretching from north-east to south-west obliquely across the Acheloos ; and, unless their two sections are to be separated from each other entirely, we must extend the Agraioi of the western bank much farther north than hitherto we imagined necessary. Unfortunately in so doing we are involved in contradiction to our best authority for the geography of this region, as it becomes impossible for the Acheloos to ' pass through ' the territory of the Amphilochoi. A further objec- tion is that, on such an arrangement, it would be hard to ^ If we are right in changing the Afpitis of Diod. xix. 67 into 'AypacK, we have an indication that this tribe made on that occasion an unsuc- cessfulattempt to advance east of the Acheloos. Perhaps, if Diodoros had told us the whole story, we might have found that the re-capture of Agrinion by the Aetolians was only the prelude to the final subjugation of the Agraeis, and that their union with Aetolia dated from 314 n. c. G a 84 THE AETOLIAN TRIBES. [ch. make Athamania border upon Amphilochia, as certainly was the case. One way out of the difficulty certainly remains, and it has sometimes been adopted. It is proposed to identify the Agraeis and the Aperantoi, either wholly or partially. No evidence for this theory exists ; it appears to spring from the notion that the modern name Agrapha is a survival of the ancient word Agraia ^. Agrapha, however, has no connexion with the Agraioi, and affords no ground for bringing them to the east of the Acheloos. The derivation given by Leake is far more probable ; the word dates from Byzantine times, when the villages were ' not written ' separately in the pub- licans' books, but the inhabitants in a body accounted for their taxes ^ The only evidence, therefore, for putting some portion of the Agraeis on the east of the Acheloos lies in the expression of Thucydides relating to the course of the river, —an expression which is not pressed in the case of Dolopia and Amphilochia ^. Why then should it be insisted upon in the case of -the Agraia ? Thus, in a circuitous manner, we have obtained more or less precisely the limits of the third and largest of the ancient Aetolian cantons; that of the Eurytanes. Starting from the chain of hills that closes the valley of the Spercheios on the west, linking together Mount Typhrestos and the range of Kallidromos and Oita, the Eurytanian frontier ran west- wards along the continuation of Mount Veluchi in the direc- tion of Stenoma and Kerasovon until it reached the confines of Aperantia, probably at the river Megdhova *. From the ' Becker, Diss. i. i8, note 78 : ' Ex voce 'Aypala interposito digammate regionis montanae, quae in sinistra Acheloi ripa est, et pagi in ea siti nomen Agrapha, quod nunc in usu est, fluxisse videtur.' ^ Leake, N. G. iv. 266. ' As a matter of fact the words of Thucydides {piav . . . bia . . . 'Aficju- Xoxav) cannot be pressed with regard, at any rate, to the Amphilochians. It has yet to be proved that the Amphilochian frontier actually came as far east as the Acheloos. For an expression precisely parallel to that of Thucydides, cf. Strabo, p. 327 : aiiTos he 6 'h^ekmos els T^y 6a.\aTTav Kai 6 EvTjvos, 6 fiev Trjv 'AKapvavlav Sie^iiiv 6 8e rrjv AlraiKiav, where the word Sie^iav, though strictly true of the Euenos, could hardly be pressed in the case of the Acheloos. * Leake, N. G. iv. 275, extends the Eurytanian frontier much farther viii] THE THESTIEIS. 85 left bank of the Agaliarlos river the territory of the Eurytanes was still conterminous with that of the Aperantoi as far south as the borders of Ancient Aetolia, and sweeping along the northern verge of the central plain their frontier embraced the eastern end of the greater lake. Then, turning north- wards along the Phidharis, it joined the boundaries of the Ophioneis, or more accurately the Bomieis, on the water- shed formed by the eastern prolongation of Arabokephalon. Whether the whole of this great area was the possession of the Eurytanes may well be doubted. The historians tell us nothing definite, and the meagre winnings of travel and conjecture do not enable us to answer the question with absolute certainty. In 218 B. c. Philip suddenly made his appearance in the heart of Aetolia. Polybios, describing his route with minute- ness and accuracy, says that the army kept Stratos, Agrinion, and the Thestieis upon its left flank ^. In the list of towns passed by the Macedonians on the right and left hand it is noticeable that we find only a single instance of the people being mentioned instead of their city. We can scarcely doubt that in the position indicated by Polybios there existed a tribe called the Thestieis, a name connected with the earliest history of the Aetolian land. Amidst all the contra- dictions that obscure the history of primitive Aetolia, a hero named Thestios always appears as a prince of the Kouretes, ruling in Pleuron ^. It is only later invention that makes him a kinsman of the contemporary Oineus, the Aetolian monarch northwards. He writes : ' It is highly probable that the crest of the ridge of Agrapha formed the ordinary boundary between Aetolia and Thessaly. In that case Mount Karava was the extreme northern point of Aetolia.' Mount Karava~s in Lat. 39° 20' N.,and fifty kilometres N.W. of Veliichi. Leake thus makes Aetolia conterminous with Athamania, as he confines the Dolopes entirely to the eastern, or Thessalian, side of the mountains which run in a S.E. direction from Mount Kariva to HSghios Ellas, south of the town RendSna. But it is clear that the Dolopes stretched westwards from this ridge, and thus intervened be- tween Athamania and Eurytania, although the lines of demarcation must be entirely conjectural. ' 1 Pol. V. 7. " Str. p. 461 : Tvapa QioTiov tov rav liKevpaviaiv &p\ovTa. Id. p. 466 : iniKpa- Telv fUvToi Qiunov rijs IlXevpaiw'ar, rbv irevBfpov tov Oivitas h\6aias 8e jrarepa, fiyoviievov t5>v 'K.ovpTjTav. See also Paus. iii. 13. 8 ; Strabo, p. 465. 86 THE AETOLIAN TRIBES. [ch. of Kalydon. In whatever way we may attempt to rationalize the myth of the Kalydonian boar hunt, that part of it which relates to the slaying of the sons of Thestios by the enraged Meleagros appears to conceal a piece of genuine history ^ The Aetolian invaders who had established themselves in Kalydon succeeded in breaking the power which they found in possession of the land. The Kouretes, or at least that section which bore the name of Thestios, were expelled from Pleuron and compelled to retire from their old abodes on the coast into the interior of the country ^. Although subse- quently the family of Oineus and the splendour of Kalydon decayed, the catasti-ophe had no effect upon the fortunes of the people of Thestios ; Agrios and his sons take their place in the legends *. The Thestieis of Polybios are the flotsam and jetsam of Heroic Aetolia. The hills on the south of Aperantia provided a refuge from the invaders who overran the plains of the coast and possessed themselves of the fair heritage of the Kouretes. It seems most probable, therefore, that the Thestieis occu- pied the well-marked triangular section lying between the river Zervas and Mount Viena. On the south-west their territory would reach as far as the plain between the Ache- loos and the modern Agrinion, thus touching the domain attached to the ancient town of that name*. The land of the Thestieis is roughly bisected by the river Eremitsas. How must we account for the variation in the language of ^ Cf. Str. p. 466 : TroKe/iov 8' ijmea-ovTOS toIs QiaTiabais npos Olvda (cat MfXe- aypoVf o)ff fi€V 6 Troirjrfjs ' a/x0i (Tvos KetjioKy Ka\ depfiari ' Kara rfjv Trept tov Kairpov fivBoKoyiav, i>s fie to cIkos nepi fiipos Trjs x&pas, k.t.X. ^ We hear an echo of this in the assertion quoted from Ephoros (Strabo, p. 463), that the Kouretes crossed the Acheloos into Akarnania under pressure from the Aetolians. See also id. p. 465. ° Paus. ii. 25. 2 : Olvca yhp TOV ^aaCKevo'aVTa iv AlraKia Xeyouo-ii/ in6 t&v 'Aypiov naibav eK^Xridevra Trjs apxrjs iraph Aw/irjSrjv is "Apyos aiKe(r8ai. 6 Si to ptv oKXa iripapriiTev aira crrpaT^vcras is ttjv KaKvSmviav, /c.t.X. Apoll. i, 8. 6 : 01 fie Aypiov iratSes . . . ac^eXd/xevot ttjv Oifias /3ao"tXeiav tw Trarpi efioffaw, kol Trpoo'eri C&vra TOV Olvea KaScip^avTes iJklCovto. See also Hyginus, Fab, 175. * In the extreme west, therefore, the Thestieis would reach as far as the Acheloos. This derives some support from the statement that the Acheloos, before it got that name, had been called Thestios (Plut. De Jluv. xxii. i). At a still earlier period it was known as the Axenos. Strabo (p. 450) tells us that the Acheloos had also once borne the name Thoas. With reference to the name Thestios, see p. 178. vin] THE THESTIEIS. 87 Polybios in enumerating the points to the right and left of the Macedonians ? The reason is that in the interval between the ancient Agrinion and the lakes there existed no town that could naturally serve the historian's purpose. The three kastra ^ in this region are all too far distant, are geo- graphically too much dissevered from the plain, to admit of being used as points of reference. It was otherwise with the territory belonging to the tribe ; that extended at least to the borders of the plain, most probably to the very margin of the lakes. On the banks of the Eremitsas itself two more tribal names have survived, graven on the rock, in situ for ever. An ancient boundary stone is still extant, proclaiming ' the limits of the Eiteaioi and the Eoitanes.' The stone gives us a fixed point in the topography ^. The river flowing past the stone was the division between the two tribes ; but we cannot say on which side of it each respectively fell. We must suppose the Eiteaioi and Eoitanes to have been two sections of the Thestieis. Thus the significant physical bisection of the area inhabited by that tribe was adopted in politics. In what precise relation the Thestieis stood to their neighbours the Eurytanes must be left undetermined ^- The analogy of the Ophieis, among whom the existence of minor cantons is an historical fact, and the discovery of the two subdivisions of the Thestieis, together with that of other tribal names of which the cantonal affinities are unknown *, suggest that similar remains may yet be found to reveal to us something certain in place of the conjectures with which we must content ourselves in reconstructing the arrangement of the Aetolian tribes. ' Those of Mavrovru, Vlochos, and Paravola. ^ See p. 180. ' Is it possible that 'Eoirai/es is the genuine Aetolian form of the name which appears in Thucydides and Strabo as UpvTaves, by the substitution of an inteUigible for a meaningless combination of syllables ? If this should prove to have been the case the views expressed in the text will require considerable modification. One of the Attic Demes bore the name EiVea. * Such as the Bouttioi, Porioi, and Phyllaioi, of the Skala inscriptions. TOPOGRAPHY AND ANTIQUITIES . . . jroW^s &v SLKaltos Tuy- Xavotfiei/ (TVyyvafiTjs ' . . . ^ ^rifi/iao'L )(pa)^poi rots avrdlsj rj )^fipiovvTes' TOVTO yap ov irapaiTOv- fixBa. Pol. xxix. 12. CHAPTER IX. The Coast Plain. I. Both on account of its position near the eastern confines of Old Aetolia, and on account of its ancient reputation, Kalydon may justly claim to be the starting-point of the topographer. The city was the theme of poetry from Homer to Statins. The Catalogue mentions ' Chalkis by the sea, and rocky Kalydon^.* In the legend related by Phoinix, Meleagros is promised a gift of land chosen 'where the plain of lovely Kalydon is fattest ^.' Euripides admirably describes the site in the words quoted by Lucian, — ' This is the land of Kalydon whose rich plain fronts the passage to Pelops' isle '.' Finally, Statius sings : — Et praeceps Calydon, et quae Jove provocat Idam Olenos *. In history the town makes no great figure. It is identified almost entirely with Heroic Aetolia. The names of Oineus, Tydeus, and Meleagros throw round it a poetical splendour that fades away in later times. Soon after the close of the Peloponnesian war Kalydon is found in the hands of the * Horn. //. ii. 640 : XaXxtSa t ayxiaXov KaXuSSvd re tTeTprjeaa-av. ' Id. II. ix. 577 : 'OimoOi viorarov TreSiov KaKvSavos ipavvris. ' Frag. Meleagr. i. (518) : — 'K.akvbav /lev iJSe ytua, UeXoirflas x^owos ev avmropdiiois ireSC txova- eiSai/jLOPa. Cf. Frag. 561 : — 'i2 ytjs Trarpcpas X"'P* (jji^TaTOV vibov T&nkvhaivos. * Theb. iv. 104. Cf. Ovid, Met. viii. 522 : Alta jacet Calydon. 92 THE COAST PLAIN. [ch. Achaians, who apparently incorporated it politically \ until the battle of Leuktra broke the Spartan power ^. Its position, close to the entrance of the Corinthian Gulf, made Kalydon the sentinel-city of Old Aetolia, guarding the approach from the side of the Peloponnese. This comes out clearly in Xenophon's account of the expedition of Agesilaos and the difficulties which threatened his retreat from Akarnania ^ In 48 b. c. the old strategic importance of the town reappears, but only for a moment*. At the hands of Augustus, Kalydon received her death-blow ; her population was transported to the new city Nikopolis, and Patrai was enriched with a share of her spoils ^ In the- days of Strabo, both Pleuron.and Kalydon, 'once the orna- ments of Greece,' had fallen into the most desolate con- dition ^. Pliny, usually a poor guide, gives us exact particulars concerning the position of the city '. He says : ' Kalydon Hes near the Euenos, seven thousand five hundred paces from the sea.' Skylax also enumerates it among the cities on the coast ^- As to the identification of the Euenos there can be no question. The modern Phidharis, the only truly Aetolian river on this coast, must be that Euenos of Aetolia ' Xen, Hist. iv. 6. I : Mfra hk tovto 01 'Axaioi txovTfs KaKvhava, rj to TroXatov AirtoXiaf ?», koI jroXiVas neiroirjiievoi, Toiis KdKvSavimis, (j}povpelii ijvayKa^ovTO iv avTJj. ^ Diod. XV. 75 : Au^iijy Koi SavnaKTOV koI KaXvSava (jipovpov/ievrjv lur' 'Axai&v rjXevBcpairev, sc. Epameinondas, in 367 b. c. ' Xen. Hist. iv. 6. 14 : 071^61 Trc^rj 6i' AJruXiar TOiavras ddovs &s oSrf ttoXXoI oVt€ oXtyot biivaivT &v olkovtcov AlrtaXap TropevetrdaL . , . eVeidi) 8e iyivero Kata to 'Pi'oc, Tavrrj Stomas oiKaBe aTr^Xtfe' Km yap tov ck KoKvSavos cKit\ovv els HcKottov- prjv iv '0<^iev(rij'. ^ Ptol. iii. 14. 12 : Toiv de iroTafiav . . . 6 ^Se EUtjvos ev t^ KaXXi8piifia> 3pet iiriarpe^wv k.tX. ' Dion. Kail. /. 61 : 7roTaix6s t 'Evrjvos « nivSov peav. * Str. p. 460. " Thuc. ii. 83 : kotciSov tovs ^\.6-qvalovs airo rrjs XaX/tiSos Koi tov Evrjvov Trora^oC TTpotrrrXeovTas crcjjiiTiv. * So Meletios, Geogr. ii. 306 : Evrjvos, Koivas ^Mpi. Gell, IHn. p. 292, calls it Ophitari ; Pococke, Descript.of the East, ii. 175, Aphidare : both by a very natural mistake. The names ^venos and Phidharis are used with about equal frequency by the peasants. ' Str. p. 451 : inaKeko 8e Avicoppas irpoTepov. Steph. Byz. AvKoppas, jrorapos, oVTivesEvrjvovcjjaai. Lykoph.^/ex 1012: Kai tov iic AvKoppaiav Trorav \ (TTparrj- Xarrfv (rvv, Kaprepov Topyris tokov, i.e. Thoas, cf. Hom. //. iv. 253, 527. Hyginus, Fab. 242 : Evenus Herculis filius in flumen Lycormam se prae- cipitavit [quod nunc Chrysorrhoas appellatur]. See also Plut. De Fluv. viii. ; ApoUod. i. 7. 8 ; Strabo, p. 327. * Explained in Etym. Mag. as irp^os, koi perpios, koi pfi TopaxwV- Con- trast Philostr. jun. Imagines xvi. : M17 bibiBi, & nal, riv ESt/vov jrorafibv ttoXXw KvpaivovTa Koi, vnep t^s Sx^"-^ alpdpevov. 9 Fqy. iii. 543. i» Cf. Leake, N. G. i. 108. 94 THE COAST PLAIN. [cH. of the outrage upon Deianeira and the death of the Cen- taur^ — °Os Tov ^aSippovv TTOTafuiv ESrjVOV ^poTovs fii(r6ov VdpeuE pfEpcriV. Upon one or other bank of the Phidharis, therefore, we must look for the site of Kalydon. Having regard to the connexion of the city with the very dawn of Aetolian history, we should naturally turn first to the west of the river; but certain words of Strabo cause us to hesitate. He tells us that Pleuron was, as it were, the capital of Plain, or Old, Aetolia ; and Kalydon of Aetolia Epiktetos, — Kalydon getting the epithets ' i-ocky ' and ' lofty ' from the mountainous dis- trict in which it lay ^. Pouqueville was led by this to look for its ruins upon the left bank of the Phidharis, in the hilly country behind Mounts Varassova and Klokova, between Aetolia and Lokris. This view, however, has no better sup- port than that given by the confusions of Strabo and the imagination of the French traveller. No Hellenic remains can be found upon the left bank of the Phidharis, with the exception of those which we are compelled to attribute to Chalkis. Pouqueville, it is true, alleges that he saw upon the slopes of Mount Varassova, above the village of Mavro- mati, fragments of an akropolis wall in the style of that of Mykenai, mingled with repairs due to a later age ^- Too many proofs exist to show that Pouqueville was the victim of his imagination ; the ' inspiration ' that guided him to his wonderful identifications sometimes led him more gravely astray. Later travellers have sought in vain for those re- mains to which he gave the distinguished name of Kalydon *. It is possible that the appearance of the rocks, which on the sides of the sandstone hills behind Varassova are split into ' Soph. Track. 559. Cf. Paus. x. 38. 2 ; Apollod. ii. 7. 6 ; Lucan, Phars. vi. 366 : Et Meleagream maculatus sanguine Nessi | Evenos Calydona secat. ^ Str. p. 460 : wrav fie ^tj ri\v KaXvSwva atTretay re Kai TreTprjecra'avj diro T^ff Xa>pas SfKTfuv' eipriTm yup on rr/v xa>pav Si^" S'f^ovfr rrfv /lev opcipfiv Ka\ eViKT^- Tov rjj KaXvSStvi npoireveinav, Ttjv jreSiaSa 8e rfj nXeupajxi. ' Vqy. iii. 540 : ' des mines cyclop6ennes avec des restaurations hel- l^niques.' Id. iv. 6: ' Cependant on me dit, et je m'en assurai dans la suite, qu'il existe encore quelques pans de mur d'une acropole semblable a celle des Mycenes, et des portions de remparts d'une epoque pos- terieure. Au-dessous de Calydon, j'apercevais le village de Mavromati.' " Cf. Bazin, Mem. p. 360. ix] THE KASTRO OF KURTAGA. 95 curiously regular forms, led him to fancy that he saw in the distance vestiges of ancient fortifications. Only upon the right bank of the river, however, can remains be discovered, and they lie in exact agreement with the information quoted concerning the situation of Kalydon. The ruins to which I refer are those generally known as the Kastro of Kurtaga, identified in 1809 by Leake with the city of Meleagros ^. ~ Half an hour above the right bank of the Phidharis three bare hills rise at the south-eastern foot of the Zygos. Their disposition is such as to enclose a narrow valley opening to the south-west in face of the traveller who approaches from the village of Bochori ^. Nothing is at first visible of the city walls upon the hills, which are covered only with coarse grass and asphodel, with here and there a patch of grain, all alike parched and burnt in the fierce summer heat. A few vines in the narrow valley alone suggest a connexion between this grey desert and the city of Oineus, where the wine-god found especial honour and worship ; but Dionysos has left his old home, and the vines are blasted by disease. Suddenly an imposing structure rising before us shows that our first impressions were mistaken. A low ridge on the left bank of the torrent that flows down the vale bears striking terrace-walls, evidently designed to support some huge building placed on the level above them. The most perfectly preserved portion, some six courses or eighteen feet high, has a length of ninety-six feet. The ridge being very steep required support on the sides towards the torrent and the mouth of the valley ; but on the other sides it sinks more gently. The depth to which the masonry extends into the hill is not known, but it is a proof of the excellence of the work to find that after the lapse of so many centuries there are no signs of bulging. The style is a fair example of ' regular Hellenic,' i. e. the courses are regular and the joints upright ; but its effect is spoilt by the poor quality of the material. The ridge sinks very gradually towards the head of the valley in which it is placed. Going down the slope we pass a second much smaller terrace, now occupied ^ N. G. i. 109 ; iii. 533 fol. " Unoxapi,. i. e. 'Yiroxap^ov, ' the village in the plain.' Cf. Leake, N. G. i. 112. 96 THE COAST PLAIN. [ciH. only by the ruined church of Saint John, but showing traces of ancient work. Advancing a few yards farther in the same direction we reach at last the city wall and the main gate. The great terrace, placed in so conspicuous a position upon the main road leading into the city, can only have been intended to support a temple, the principal sacred edifice of Kalydon. Corresponding to it, upon the smaller terrace, there was probably a second temple. Our minds at once revert to the passage of Pausanias wherein he mentions the chryselephantine statue, representing Artemis Laphria, which he saw in the akropolis of Patrai^. It was originally the possession of Kalydon. Strabo, however, does not mention Artemis, but only Apollo Laphraios, whose temple, he ex- pressly says, was ' near ' the city, not in it ^. It was no un- usual thing for temples to be placed in the open country outside the walls, as they were sufficiently protected by their sanctity from outrage by an enemy. In the present case the temple terrace was virtually an outwork protecting the road and gate. There is no doubt that we must consider the cultus to have been one of Apollo and Artemis conjointly, and we can have no hesitation in localising the worship upon the great terrace. Yet neither on nor around it do we find the smallest fragment of architecture ; nothing has escaped destruction, except the foundation-blocks of the temple pave- ment on the top of the ridge. The peasants have no tradition of anything having existed in the shape of columns or such like. It would seem that the structure of the temple was much more simple than we should have expected from the value of the statue that it contained. Still, there is nothing strange in this: we are apt to forget that the temples of Athens, Aigina, and other great towns, must not be regarded as examples of the normal form, but rather as specimens of its final elaboration when translated into stone. The description of the ceremonies of the cultus, as given ' Paus. vii. l8. 9 : Harpevin te 6 AxlyovaTos aWa re tS>v ex KaKvSavos \a(j)vpav Koi dr) Koi T^j Aatjjpias eSooice to ayoKjia, o 81} Koi is e/ie en iv rfj aKpon-oXei rfj Tlarpecov (Ixf rijiiaj, ^ Str. p. 459: nepl be Triv'KdKvb&vi eVti to tov ha^piov 'AffdXXfflTOS Up6v. Perhaps this is the source of what we read in Meletios, Geogr. ii. 306 : nXijcr/oi/ rrjs KaKviavos fiTov to 'Upov tov navTOKparopos Ai6s. Yet above he writes : K.a\t]8a)v, Sirov eVejSeTo fj Aai^pla 'Pi.pTep.is, air avT^s iiCKridri 6 KaXuScii'ios Apv/ids, Kol 6 ' Aypioxoipos. ix] THE LAPHRIAN FESTIVAL. 97 by Pausanias \ applies primarily to Patrai ; but its main features must be Aetolian, for with the temple-image the whole of the sacred apparatus and the ordinances connected with it would undoubtedly be transported to its new home. This is hinted at in another place by Pausanias himself, when he says that the Messenians of Naupaktos also adopted the worship of the Laphrian Ai-temis, so that a statue of the goddess was to be found in Messene ^. There is an Aetolian character about the Laphrian Festival at Patrai that bears out this theory of its origin. ' Round the altar they erect a circular barrier of green wood, sixteen cubits high. The driest wood they can get lies within, updn thfe altar. And for the season of the feast they contrive a smobth ascent to the altar, by putting earth upon its steps. First, then, there is a splendid procession in honour of Artemis, in which the virgin priestess rides last in a chariot yoked with stags *- On the following day the sacrifice takes place ; the greatest enthusiasm being displayed, both by the public functionaries and by private persons. They throw alive upon the altar edible birds and victims of all kinds, wild boars, stags, and does ; some throw upon it wolves' and bears' cubs ; others even full-grown animals. They also put uporl the altar the fruit of cultivated trees * ; then they set the pile on fire. Thereupon I have seen a bear or other animal, under the first violence; of the flamfe, trying to force its way out, and sometimes even escaping by main strength ; but those who cast them in bring them back again to the pyre, and they say that no one was ever hurt by the animals.' This horrible holocaust was offered annually to the goddess. It is to be noticed that the niain gateway of Kalydon lies on the axis of the great terrace ; it was through this gate that the magnifi- cent procession would march up the gentle slope to the temple of the twin deities. ' Paus. vii. 18. II fol. '■' Id. iv. 31. 7. He adds : to /lev 8^ t^s Aa^pias dtpUeTO ivofia ts xe Meo-ffi;- v'lovs KOI es -TlaTpiLS 'hxaiS>v fiovovs. ' Cf. copper coin of M. Aur. figured in Num. Comnt. I. c. ; and of Elagabalus, Mionn. ii. 197, 364. They show a priestess in a chariot drawn by two stags. * Cf. Horn. //. ix. 534 ; Artemis sends the boar :— Xoia-a/jUvrj 01 ou ti BoKvaui yowo) dXu^r Olveiis pe$'. H 98 THE COAST PLAIN. [ch. The evidence of Numismatics supplements the infor- mation supphed by Pausanias. ' The figure of Artemis Laphria on coins is almost unvaried ; the only marked variation being that the bow rests in some cases on a high pedestal, in some cases on a low pedestal, in some cases on the ground. The goddess stands, her head sUghtly turned to her left, clad in a short chiton with diplois which leaves the right breast bare, a chlamys hanging over her left shoulder, high cothurni on her feet. Her hair is in a knot at the back ; a quiver is at her shoulder. Her attitude is one of ease, yet not quite free from stiffness ; the left knee slightly advanced, the right hand resting on her side ; in the left hand a bow \ The type is clearly a copy of the cultus-statue of Artemis Laphria; this is even proved to demonstration by a coin where it appears side by side with the Aphrodite of the Corinthian akropolis"- We thus arrive at an interesting result. It is distinctly stated by Pau- sanias that the cultus-image at Patrai was the work of Menaichmos and Soidas of Naupaktos '. On this Brunn remarks that its date must be earlier than the settlement of Naupaktos by the Messenians at the beginning of the Peloponnesian war*. And Pausanias says that the sculptors must have lived not much after the archaic sculptors Kallon of Aigina and Kanachos of Sikyon. But the statement of Pausanias seems exaggerated in. view of the style of the figure on the coins, which may perhaps be assigned to the middle of the fifth century, but can with difficulty be given to an earUer date. In any case this will be one of the earliest statues which represent Artemis in Amazonian form °.' The small temple that apparently stood near the site of the present church of Saint John may have been dedicated to Dionysos ; we are unable to say anything of the Aetolian form of the cult. Passing to the remains of the city, we reach first the main gate in the southern wall running across the gully. Nothing but the ground-plan remains, but that is quite clear. We have an opening seventeen feet wide and thirty-two feet long ; on the right and left a square tower springs forward eleven feet from the face of the wall to protect the passage. There is nothing elaborate about this, nor about the plan of the smaller gate- ^ This Laphrian Artemis is clearly only Atalanta. Cf Eurip. Frag. 531 : 'Apuas 'ATaXdvTi) Kvvas | Km t6^' c^ouo-a. Phoin. I162 : Tg KaXXireS^o) /ai/rpc McuvaKov Kopjj. With her arrow Atalanta draws first blood in the boar hunt. ' A copper coin of Commodus. ' Pans. vii. 18. 10. * Gesch. der Gr. Kunst. 2nd ed. i. 80 = ist ed. i. 112. " Numismatic Commentary on Pausanias (Imhoof-Blumer and Percy Gardner), p. 77. ' cn O g 3 o o 1-1 z o o O o ix] THE FORTIFICATIONS OF KALYDON. 99 ways. There are five of these subordinate entrances ; one in the north, another in the west, and a third exactly opposite in the east. The modern pathway traversing the site runs from the last-mentioned gate to the great gate in the southern wall. A fourth postern is near the south-eastern angle of the enceinte ; lastly, there is a gateway between it and the main entrance. Extreme simplicity of design characterizes the Kalydonian gates, which are, besides, remarkably few in number, — only six in a circuit of two miles. Both points are indications of an early date. To the primitive engineer, to breach his wall for the purpose of ingress and egress was to create weakness, which he knew not how to overcome without a great expenditure of material, if he attempted the task at all. The advance of the science gradually made the gateway the strongest point in the whole circuit of the defences. The conclusion drawn from a consideration of the gate- ways of Kalydon is pressed more strongly upon us by a survey of the whole enceinte, which, in a very unequal state of preservation, but nowhere more than seven courses high, can be traced throughout its entire circuit of more than two miles ^. The general plan is that of a rough quadrilateral placed across the spur of the Zygos in such a way that the wall runs along each face and crosses the ridge at head and foot, but without including the two extremities. The nar- rowest part of the enclosure is that which fronts the river Phidharis, looking towards the south-east. On the north of the site we find the rectangular akropolis height abutting upon the main wall, being cut off from the rest of the city by a short cross-wall. The akropolis looks upon a narrow vale, down which a stream flows eastwards into the Phidharis. The lines consist of an irregular alternation of square towers and salient angles. Curious variations are observable, both in the distribution of the angles and towers, and in the masonry of the fortifications. In some parts the wall is built in fairly ' regular ' style, that of the great terrace ; in others it has a much ruder and more ancient appearance, — the irregularity of the courses, their oblique joints, the unworked faces of the blocks, all combine with the brown, easily ' Leake (N. G. iii. 535), ' near two miles and a half.' Bazin (Mem. p. 356), 'plus de quatre kilometres.' H a lOO THE COAST PLAIN. [ch. weathered sandstone, of which the wall is throughout con- structed, to give the work a semblance of higher antiquity than many examples showing the same features of technique. Nevertheless, we must be on our guard against exaggerating the amount of this difference in the style of the masonry. That it is to a certain extent of an accidental nature may be proved by contrasting the walls on the northern with those on the southern and seaward face of the hills, where the corrosive action of the salt breeze has had free play upon the surface of the masonry ^. Of relics within the walls of Kalydon it is impossible to say much. A few terraces in the akropolis and on the west of the site ; substructures of fine character near the main gate ; traces of an aqueduct draining the interior of the enclosure into the gully that descends frofn the Zygos along the western wall,— this is the meagre list of the re- mains of the greatest city of Aetolia. The site has yielded a few inscriptions ^ ; one of them, on a stele found near the church of St. John, confirms the identification with Kalydon. According to Cousin ^ it reads : — 'A itoXls] Ka\(y^Sa>vca>v Adiiap)(pv .... mvoi KaXvSmvLOV top avT\ds evepyerdVi Kalydon appears in Federal inscriptions throughout the history of the League ; generally, however, merely as desig- nating the native place of Aetolian magistrates. In looking through the list of Aetolian Strategoi we find the names of Alexander and of Damokritos, two Kalydonians who more ^ If any one is sceptical as to the reality of the problems suggested by an examination of the style of masonry and methods of fortification employed in the Hellenic palaiakastra, let him contrast the walls of Kalydon, an Homeric city, with those of Old Pleuron or Chalkis, both of which, so far as the literaty evidence goes, bfelong to the same epoch as Kalydon. '^ Leake (N. G. i. 112) notices one found in some vineyards near Kur- taga, reading *IA0YMENA | antimaxos. " In the Bull, de Corr. Hell. x. (1886), p. 185. PubUshed previously by Baz. Mem. p. 361, and App. No. 13. For another Kalydonian inscription, see Bullettino, 1849, p. 105 = Ditten. Syll. 258 = Lebas. ii. 1031 = Bull, de Corr, Hell. x. p. 183. IX] POSITION OF KALYDON IN THE LEAGUE. lox than once held the highest office of the nation ^ The im- portant inscription concerning the frontiers of Melitaia and the Pereis in Thessaly incidentally bears witness to the importance of Kalydon among the Aetolian cities. A copy of the decree is ordered to be placed 'in Melitaia, in Delphi, in Kalydon, and in Thermon I' Melitaia, the town chiefly interested in the award, would naturally have a copy. Delphi, between 290 and 190 b. c, was the ecclesiastical capital of the League*. Aetolia proper is represented by Thermon and Kalydon, — the former the new political capital, the latter the old ; the one identified with Aetolia Epiktetos, the other with Aetolia Antiqua *. ' Alexander was Strategos sometime between 208-2CX); again in 196/5 ; and also in 185/4 b. c. Damokritos in 200/199, ^id 193/2 b. c. Cf. Wescher-Foucart, and Baunack {Sammlung-Collitz), passim ; Bull, de Corr. Hell. y. (1881), p. 409. In Pauly-Wissowa's Realencycl. (New Ed. Art. Alexander, No. 32) Alexander of Kalydon is -identified with the Aetolian orator of that name, who was also called d'lorior (cf. Pol. xviii. 3). But this is surely mistaken. 'l' rjs to kvvos irrjua iv KaXvfimvi, * Paus. vii. 21. I fol. ' Ibid. : care(Ta^e re avrrjv is rijv miyriv, fj iv KaKvSavi ia-Tiv ov iroppa Toi \ilievos, Koi air iKtivrjs 01 entira SvBpajroi KdKKiporjv ttjv wrj-yflv KoKovai. * Str. p. 460 : eo-Ti Si Tis Kal irpbs rrj KaXuSSw Xlfivrj iicyaXrj Km eCoxlAor, fjV fXovaiv Of iv narpais 'Pto/xaioi. It must be to this that Alkiphron, Epist. i. 18, refers, m the words Kal yivrjTal a-oi to t^s -^aXTpias KaTaymyiov KaXvSavios koKttos rj Tvpp-qviiiov iriXayos, — unless the sea between the mouth of the Acheloos and the entrance of the Corinthian Gulf was known as the Kalydonian gulf IX] KALLIRHOE. 103 its fisheries were farmed by a societas publicanorum of Patrai, and Archestratos, a gastronomic poet quoted by Athenaios, says that they produced the fish labrax in great perfection'. There is also a reference in a fragment of Nikander to a lagoon or lake which he calls Onthis, on the road to Nau- paktos, and near a certain ' lofty" hill ^' This hill is left unnamed : it might with great probability be identified with Mount Varassova'^ Whether the A-iV^jj of Strabo and the Xtju.jji' of Pausanias are identical, wholly or partially, is uncertain : and the Xliiv-q of Nikander only enhances our perplexities. We may argue* that Kallirhoe is the stream already menjioned as flowing along the foot of the great terrace, past the site to which we have conjecturally assigned the temple of Dionysos (who appears in the legend). Upon this it jnust be remarked that the words of Pausanias, ' the source not far from the harbour,' would more naturally have been used of a stream rising near the sea, than of one which runs down from the Zygos, as is the case with this water- course. Still, as Pausanias had no personal knowledge of the locality, we cannot press his words. A more decisive argument against this identification of Kallirhoe is that it leads us no farther. If it were true, then by following the direction of the stream we should issue upon the coast some- where near the ancient harbour spoken of by Pausanias. In reality, however, the stream does not fall into the sea at all, but is diverted into various channels, natural and artificial, and lost in the low ground about Bochori long before the coast-line is reached. A more probable conjecture is that the ancient Kallirhoe survives in the springs that rise at the very foot of Mount Varassova, close to the modern jetty of Kryoneri. Not far from the shore at that point remains are said to be visible in ' Athen. vii. 311 a : — niorepoL 8' erepoi noWoi \La\vSSivi re kXco'i; 'Afi^paKia T. ^ Schol. iti Nik. Theriac. 215 : 'PvTralov . . ."Ea-n Se t^j AlraXias^m tiiKavSpos Ttepi Tivav els AiTaKiaii ipxoplvav biryyovpcvos' 81' alireivr)V re Ka\i>vt]v Olavov 'PuTTijs re wayov Kai, 'OvdiSa \ipvr)v 'Stcixovto tia{maK.TOv,.is 'AntjitSviiriv re iriXa^ov. ' So Palmer, Gr. Descr. p. 498, who conjectures 'Ovdis for ns in the passage quoted from Strabo, p. 460. * As does Becker, Diss. ii. 27. I04 THE COAST PLAIN. [ch. the sea: they may possibly be those of an ancient quay^. If, however, we take them to indicate the place of the Kalydonian harbour, we must remark that the broad and treacherous Phidharis now intervenes between it and the city in a way that is quite intolerable. In order to escape from this objection we may invoke the deus ex machina of the topo- grapher, and suppose the whole tract south of a line drawn from the foot of the hills of Kalydon to the point of Mount Varassova to be of very recent origin. If the plain, including that arm of it which extends between the shoulder of Mount Varassova and the foot-hills of the Zygos, north of the railway bridge, was an estuary during the sixth century before our era ^, it is clear that Kalydon must have enjoyed direct communication with the harbour at the mouth of the river, and thus the difficulty raised by the intervention of the Phidharis would be obviated. This supposition, however enticing, cannot be accepted. For it involves the assumption that during the last 2500 years the river has thrust the coast-line at least seven kilo- metres to the south ; which gives to the plain a mean annual increase of three yards ^ The rate must be set down as much greater if we take the springs of Kryoneri to be indeed the Kallirhoe of Pausanias ; for then we are committed to the acknowledgement that in his time (180 a. d.) those springs were already visible*, — that the coast-line had practically ^ Sep the Adpiiralty Chart (Cap. Mansell, 1865), No. 1676. It is cor- rected to 1894. ?^^ P- '^^• " If we are to argue about the matter at all, we cannot consent to refer the story to an earlier date. ' Although not large in itself, this amount is too great if Neumann is correct in saying that thp Spercheios, since the days of Herodotos, has only succeeded in pushing the coast-line outwards to the extent of some eight or twelve kilornetres. For the Spercheios is a much larger stream than the Phidharis, and drains a much larger tract of easily disintegrated soil. Cf. Neum. u. Partsch, Die Physikalische Geogr. v. Griech. p. 350. The rate in the case of the Phidharis must be still further augmented when we take into consideration the fact (known from Pliny, H. N. iv. 3) hat, already in 79 a. d., Kalydon was eleven kilometres from the sea, although PUny undoubtedly exaggerated the distance. As he is probably reproducing the statement of an older writer, we shall have to push yet farther back the date at which the accumulation of alluvium had ceased. ' As indeed follows from the actual words of Pausanias : t^v mjyriv, ^ ev KaXwdSvi ctrriv, . . Trjv irriyriv KoKovtri. IX] THE KALYDONIAN HARBOUR. 105 attained its present shape. And, in fact, for some centuries before the time of Pausanias the changes effected by the river, including the silting-up of the ancient harbour, must have been nearly complete, seeing that the geographical writers never hint at the existence of a harbour below Kalydon. The alluvium, long before 180 a. d., had reached the limits of its extension outwards into the gulf of Patras, and further deposits were swept away (as they are now) by the currents. Similarly, at the extreme western end of the alluvial belt, near the mouth of the Acheloos, the increase of land appears long ago to haive ceased ^. Such considerations must convince us of - the falsity of the hypothesis that the entire tract of level ground below Kalydon is of recent growth. Its creation must, on the contrary, have been already an accomplished fact at least as early as the fifth century , before our era. The Phidharis must then have entered the sea through a delta pushed out, as now, far in advance of the original shore-line. On each side of this projection there would be a bight or bay. That on the east, at the base of Mount Varassova, is now silted up : that on the west has not yet entirely disappeared, but remains as the lagoon of Bochori. We end, therefore, as we began, — with the difficulty that if the Kalydonian port lay at the foot of Mount Varassova, at the point indicated by the springs of Kryoneri and the submarine remains, it was separated from the town by the river Phidharis. The question thus resolves itself into an estimate of the value of our authority, Pausanias. His account does not reproduce the actual state of things existing in his day. If the harbour which he mentions was that at the base of Mount Varassova, it had disappeared long before his time - Cf. the prophecy of Thucydides (ii. 102) with regard to the Echi- nades : koI eio-i tSv vfjo-mv al rpreipavrai, iXiiis 8i koI jrda-as oin ev ttoXXm nvi av Xpova TovTo 7rn6iiv. Pausanias (viii.24. 11) is much put to it to account for its non-fulfilment : ras Se 'Exivd8as vfjo-ovs iiro tov 'AxeXaou fiij (T0as ^irfipov axpi r]iia>v arreipydcrdai ■yeyove 8i' alrias to AiTtoXav Wvos' k.tX. — the depopulation of the country diminished the quantity of mud carried down by the river ! It is perhaps not too much to say that the coast-line about the Acheloos mouth has been stationary precisely during the period to which the chief deposits of the Phidharis are assigned by the theory which we are com- bating: yet the land-creating power of the Acheloos must greatly exceed that of the Phfdharis. io6 THE COAST PLAIN. [ch. and had become a mere landing-place for boats, just as it is to-day : if it was the bay on the west of the river, it was then silted up and reduced to the condition of a lagoon. Pausanias gives us a legend, the details of which bear but little reference to realities. We reach, then, the following result. The Kallirhoe springs are those near the khans of Kryoneri at the base of the rock. Here, during the fifth century, there was a small port, the fast disappearing bight on the east of the Phidharis delta. A reference to this port is to be found in the passage of Thucydides in which he speaks of the Athenians having their ships stationed at the Aetolian Chalkis and the mouth of the Euenos ^. This also was the harbour referred to by Pausanias, although in his day it was no longer in - existence. The harbour used by the early inhabitants of Kalydon must have been the bight upon the western side of the delta. The stormy times of later Aetolian history w6re fatal to the economy of the country ; and the gradual decay of the great towns of the coast, culminating in the deliberate depopulation of the whole land, allowed natural forces to work unchecked. The Kalydonian port was partially silted up, and became the lagoon Onthis of Nikander, the ' great lake ' which Strabo describes as a source of wealth to Roman speculators at Patrai. Its salt- works and fisheries are still valuable ^- CHALKIS. Ten minutes' walk from the ford of the Euenos below Kalydon brings us to the rock of Varassova ^ The railway runs down between the rock and the river to Kryoneri, the starting-point of the system of North- West Greece*. Kryo- neri, officially called Kalydon, consists only of three or four rude khans at the base of the rock ; its name is derived from ^ Thuc. ii. 83. See p. 93, note 5. 2 Cf. Trik. 'I(TTopia. ii. 364, who says of the whole lagoon : jreptex« 8vo SXiKas, Trji/ iiev Kara to 'AvaroXiKov Trji/ 'Aa-7rpi]v, ttjv 8e Kara to MiroxSipi Triv MavprjV. = An alternative name is Moiint Galata. Cf. Leake, N. G. i. 107. * Kpvovepiov, Kpvovepiov "Op/ios. From this point we cross in less than one hour by steamer to Patras. IX] CHALKIS. 107 the copious springs of the purest water gushing from the foot of the mountain close to the beach : one of them is actually in the sea, and freshens the surrounding wat^r. The immense rock, rising almost perpendicularly to the summit, effectually bars all farther progress in this direction. The only path runs round the northern shoulder of the hill, where other springs are found, then along the valley of the Phidharis, there flowing from east to west, and so down the narrow but beautiful vale of Gavrohmni, between Varas- sova and Klokova. From Gavrolimni, the Kaki-skala^, — a road cut in the precipitous face of Mount Klokova, — affords communication with Rumilia and the town of Naupaktos. Varassova is the Mount Chalkis of Strabo. 'After the Euenos,' he says, ' comes Mount Chalkis, called Chalkia by Artemidoros ^.' He falls into confusion in attempting to reconcile his two authorities, Artemidoros and Apollodoros. The former placed his Mount Chalkia between Pleuron and the Acheloos ^ ; but the latter insisted that it stood above Molykria, which was east of the Euenos, and put Kalydon between Chalkis and Pleuron. Strabo, therefore, gratuit- ously imagines two mountains, — Chalkia near Pleuron, and Chalkis on the east of the Euenos, corresponding to the modern Varassova *. The town arid the mountain bore the same name. As a city, Chalkis 'by the sea' is as old as the Homeric Catalogue^. To Thucydides also Chalkis is a maritime town, as is clear from the passage already quoted, in which it is conjoined with the mouth of the Euenos as marking the point from which the Athenians bear down upon the Corinthian fleet ^. Its exposed situation contributed to its capture by the Athenian Tolmides in 455 b. c, some thirty years before the naval victory of Phormion ; but even before that date the city had been wrested from the Aetolians by ^ Ka/cij SkoXo, ^Koka MavpofiiiaTr]. Mt. Klokova itself is often called Kakl-skala ; even in Leake's Travels the name Klokova does not appear. ^ Strabo, p. 459 : /nera 8e rbv Eir/vov to 8pos ij XoKkIs, rjv XaXKiav npr/Ktv ApreiilSapos. ' Id. p. 460 : fiera^i rev 'A;\;eXo)OU Koi rrjs IlXevpavos iSpvaiv avrtjv. * Str. /. C. : el fifj &pa erepov dereov to Trpos nXevpwvi Spos XaKniav naXoipLcvov, erepou Si rrjv XaXxiSa tijv irpos MoKvKpeiq, ° Horn. //. ii. 640 : XaKKiSa t ayx'aXoi'. ^ See pp. 93, 106. io8 THE COAST PLAIN. [ch. the Corinthians \ As in the case of Kalydon, the difficulty with respect to Chalkis is not one of topography. The evidence of the ruins themselves is clear and decisive. We have one name Chalkis, and a single site : the two must go together. If the passages of Thucydides alone remained they would be sufficient to establish the identification ^. From the picturesque khan of Gavrolimni it is two miles down to the sea below the hamlet of Vasiliki *. Then, after crossing a low rocky ridge in front of Var^ssova, it is a short but toilsome climb up the mountain to the northern wall of the kastro. This wall stretches across the path, and when it was perfect it completely closed the road to an enemy advancing from this quarter. Bazin gives the name of the kastro as Pangkali*; Leake as Ovriokastro ''j a title almost as common as the conventional Palaiokastro. The popular belief is that the remains are those of Kalydon, a name which to the peasants is in no way associated with the ruins above Kurtag^. It is a fine specimen of ancient fortification that meets the eye, mutilated in some part^ even to complete disappearance, in others almost perfect. The form of the enclosure is peculiar, for it is nothing more than two practically straight walls, one towards the north, the other towards the south ; the former closing the road from the vale, the latter that frpm the sea. On the west the precipices of Varassova, qn the east unscalable cliffs, connect the two lines, and at the same time make walls i^nnepessary on those ' Thuc. i. io8 : XoXkiSo KopivBiav TrdXtr eTXov. In the same expedition he captured Naupaktos, €^ ivos, ' below,' i. e. south of, Kalydon. This is only to be explained by reference to that passage in which he inaccurately placesj Kalydon upon the left bank of the river Phidharis, and therefore somewhere at the head of the vale of Gavrolimni (Str. p. 451. See p. 132). The true position once given to Kalydon, it is obvious that one would not naturally define the situation of Chalkis by reference to Kalydon at all; some other point of comparison would suggest itself. ' TavpoKiiivrj. Ba(ri\iic^. * UayKaX^. Mem. p. 362. ° N. G. i. III. My own guide called them Z/ot-tj or Seon;. Lolling (Iwan Mailer's Hand. iii. 139) puts Elaos 'vielleicht auf dem hohen ovalen Hugel nordlich von Sesti, 40 Min. von Kalydon am Wege nach Mesolongi.' Thus two places bear the name, or my guide was mistaken in applying it to the kastro of Gavrolimni. Lolling's S^sti must be the site that we identify as Halikyrna. IX] STYLE OF THE WALLS. 109 two sides. The system of defence is that of short curtains and square towers. The northern wall is a good specimen of a pecuHar species of masonry ; the blocks, with a rudely dressed face, are disposed in regular courses, but the joints are obhque. This style can scarcely be designated 'regular Hellenic,' nor yet ife it 'irregular' properly so called. The regularity of the courses is not quite perfect ; here and there we find stones with portions cut away, the blocks being afterwards squared up by the insertion of smaller pieces very accurately fitted : the symmetry of the courses is thus marred and the appearance of true ' irregular Hellenic ' work is produced. The style of the masonry and the elaboration of the details are far ahead of anj^hing seen in Kalydbn, and do much to disturb our notions of what is appropriate to a city as old as Homer. In the ruins before us we readily imagine that we have an example of the highest development of the art of fortification in Aetolia. The most interesting features of the wall are the towers, one of which, that at the extreme north-western angle of the lines, is preserved to a height of ten courses. What strikes us at once is the manner in which they are built as integral parts of the wall and inseparable from it, not as simple projections affixed as it were to the wall after it was completed. The result is that, whereas in the latter case the terre-plein of the tower is on a level with the top of the wall, at Chalkis it is at the natural interior ground level at the base of the wall. A further result must be noticed. When the tower is merely affixed to the wall, it is not provided with loopholes, unless indeed it rises far higher than is really necessary. On the other hand, in the mode adopted at Chalkis, the tower does not project far above the crest of the parapet, and is perforce furnished with loopholes in face and flanks. Lastly, it is evident that at Chalkis we shall look in vain for steps or other device to enable the defenders to man the walls and towers ; the garrison went into the towers straight from the general level of the ground inside the walls. The towers of Chalkis, in fact, needed only three sides ; the fourth side, that towards the town, and farthest from the enemy, not serving any purpose of defence. What we actually find on this fourth side is a single wall no THE COAST PLAIN. [ch, admitting of the passage of the men into the tower past one or both ends. In the fairly preserved tower in the northern wall, just to the right of the point at which we enter the enceinte, we see that the upper flank has been produced inwards to join this fourth, or interior wall, while the other flank has stopped short and so allowed passage into the tower. In the fine example before alluded to, at the upper end of the northern wall, both flanks have stopped short, so that we have two doorways. The lintel of one of them, composed of a single block, is still in situ. The curtain is built in the usual style, with an outer and an inner face, bonded together at intervals by means of cross-pieces, and filled with earth and boulders ; on the other hand, the walls of the towers have only a single stone in the thickness. Near the point at which our path strikes the line of the northern wall the 'filling' has disappeared from between the two faces, so that the skeleton of the wall is left, showing very clearly the mode of its construction. We notice two entrances in this northern wall, one of them just below the larger tower. They are, respectively, just under and just over five feet ; a single stone forms the lintel. The wall on the south is concave, with the concavity facing the sea. It presents no additional points of interest, with the exception of the main gateway piercing its centre. The opening has a width of eleven feet, and it is defended by a tower. The strength of the gateway is very great, as it is completely flanked by the fire of the defenders stationed along the two curving arms of the main wall. The nature of the ground was probably here the prime cause of the adoption of this eff'ective plan ; but we shall find it employed elsewhere in Aetolia. In our first moments of admiration we expressed the opinion that the ruins of Chalkis represent the acme of Aetolian military engineering. It seemed impossible to put Chalkis upon the same level as the other Homeric cities. Is this opinion justified, or have we discovered anything to cause us to attribute to the remains a rank lower than at first sight seemed their due ? Our answer to this question must be that the apparent superiority of these ruins over those of Kalydon is largely an illusion. It is a superiority, not of science, which would indicate a later date than that of « H Pi O M K H b O o 5 u f- z u K H s O < IE O o ■J t Missing Page Missing Page ix] HYPOCHALKIS. 113 Strabo writes:— 'Taphiassos and Chalkis, mountains of Aetolia, of some elevation ; upon them are the towns Makynia and Chalkis, which latter is named after the mountain, and is also called Hypochalkis ^' Chalkis the town, therefore, was also called Hypochalkis, from its situation at the base of Chalkis the hill. There is no ground for the difficulties of the earlier writers on Aetolian topography, such as Kruse and Becker. The former regards Chalkis as the citadel, and Hypochalkis as the town in the plain 2. Becker thinks that Hypochalkis was the small port (which is now a marsh, called kiixv-r)) at the foot of the kastro which we have de- scribed ^. This, he says, was the site of the Chalkis of Homer . and Thucydides, the ' loniis et fluctibus hospita portu Chalcis ' of Statins*. In course of time, owing perhaps to the Corinthian and Athenian conquests, or the inroads of the Macedonians and their allies, the town was, he imagines, removed to the heacj of the valley, about the village of ' Mavromati. The X(£A.(ceia of Polybios^ is thought by Becker to be the name of the valley itself, between the port and the new site of the town. All this is very clumsy and gratuitous. There are no ruins at Mavromati; and Chalkeia, like the Chalkia of Artemidoros, is a simple variant for Chalkis. Becker seems to have been led to his hypotheses by Ptolemy, who makes Chalkis an inland town^- The words of our English historian of ancient geography are a sufficient criticism on this : — ' The blind, and almost superstitious, reverence, with which Ptolemy was regarded throughout the Middle Ages, has descended in some degree to our own days : and it is not uncommon to find writers referring to his statements, as if his apparently definite and scientific results must necessarily be based upon definite information and scientific calculation. . . . Even at the present day there still remains a lingering desire to prove him in the right if possible, and to believe in the accuracy of ^ Str. p."45I : Ta^iatro-ov rat XaXxi'So, Spr] iKavas ^rrjka, iv AlraKav neSiov p4ya' tou 8ia nerrtrov (rvperai oKkov ayav 'A)(€\aios dpyvpoSivrjs. Cf the Latin translations of Avienus (/, 591) and Priscian (/. 445). * Str. p. 460. x] THE TEMPLE OF ATHENA. 125 Kerasovon and Pappadhatais, suits Strabo's words well enough. Although he ought to be speaking of the old town, the poet Statius gives us a description of the view from the new foundation, the only Pleuron with which he would be acquainted. Tydeus cries to Athena : — ' If I return to war- like Pleuron I will raise to thee upon the heights in the midst of the city a temple rich with the sheen of gold; thence shall one look with deUght upon the storms' of the - Ionian main and the place where the turbid Acheloos, bursting through the barrier Echinades, lashes the deep with his tawny billows \' The description is nearer the truth than we might expect ; it is, in fact, a fair picture of the view to the west from the kastro of Kyria Eirini, — the opening of the Gulf of Corinth into the Ionian Sea and the promontory of Kurtzolari which covers the mouth of the Asprop6tamo,~and there are no other remains upon the Zygos from which such a view can be obtained. Finally, if Thucydides enumerated the places in their correct order, his account of the march of Eurylochos from Lokris con- firms the identification. The Spartan general moved from Molykreion to Kalydon, Pleuron, and Proschion, where he remained until he crossed the Acheloos ^- The temple of Athena at Pleuron seems to have had some celebrity. Dionysios, son of Kalliphon, writes^: — 'Next comes Aetolia, in which is Pleuron with the holy shrine that bears Athena's name.' The reference on the part of Statius has already been quoted. Leake suggested that the By- zantine church on the akropolis of Kyreirini might indicate the site of this temple : but we must ask the question, to which town did the shrine really belong, — to New or to Old Pleuron ? If the poetical Description of Greece which we have quoted under the name of Dionysios, son of Kalliphon, were really the work of Dikaiarchos, to^ whom it used to be ascribed, or if it was based upon the r^s Yleplotos of Dikaiarchos, it would afford a presumption that the temple belonged to the older city ; for Dikaiarchos was 1 Stat. Theb. 727 fol. ^ Thuc. iii. 102. 2 Dion. Kail. /. 57,:— "Exerai 8' tdroKia, iv jj TToAiy UXevpav VTroKfirm, X'.ep6v S.yiov 'Adtjvas etrriv avo/juKTiisvov. 126 THE COAST PLAIN. [ch. a contemporary of Alexander the Great, whereas New Pleuron is certainly posterior to 235 b. c. We may perhaps take it, then, that the temple belonged to the old city, and that it continued to exist after the lower site had been abandoned by its inhabitants in favour of the more secure slopes of Mount Arakynthos. If then Leake is right, as he probably is, in locating the temple of Athena upon the akropolis of Kyra-Eirini, we have two alternatives before us, — either the shrine existed from the first upon the summit that, sub- sequently to 235 B. c, became the akropolis of New Pleuron, or one was there built to take the place of the original shrine on the site of Old Pleuron. The former is perhaps the more probable hypothesis ^. If we accept it, the description furnished by Statins turns out somewhat curiously not to be the anachronism we imagined, for the Ionian Sea and the estuary of the Acheloos were, and always had been, visible from the temple on the hill-top, the building of which lay far back in a forgotten past. With respect to this Aetolian cult of Athena we are unable to say anything. Aetolian gold coins show the head of Pallas with Corinthian helmet ornamented with crest and griffins^. Though there is nothing distinctively Aetolian about the type, we may perhaps see in it the contribution of Pleuron to the national coinage ; just as the jaw-bone and head of the boar were derived from Kalydon. The same principle underlies the selection of deities to represent Aetolia at Delphi, where, by way of commemorating the defeat of the Galatai, the League dedicated statues of Athena, Artemis, and Apollo*; in which we must undoubtedly see a reference to the worship of those deities at Pleuron, Kalydon, and Thermon respectively. The unfortunate fact that the Aetolian cities issued no separate coinage, and that consequently the Federal coins date only from about 280 b. c, ^ In that case the temple must have been outside the walls, so long as Old Pleuron in the plain continued to be inhabited. Cf. that of Apollo and Artemis at Kalydon, p. 96. ^ Brit. Mus. Cat. Thessaly to Aetolia, p. 194. ' Paus. X. 15. 2 : 'ApTf/uSos, tA Sc 'ABrjvas, Svo tc 'AiroXKavos aydXjuara etrnv AItoKSiv, riviKa a-^'uriv i^eipyairBri rh. is VciKarai. TwO Statues of Apollo, to mark, perhaps, the superior position of Thermon : or on account of the close connexion of Apollo with the whole story of the invasion and its repulse. x] HISTORICAL POSITION OF NEW PLEURON. 127 prevents our gathering anything from Numismatics with regard to local types of divinities in AetoUa ^ Whatever may be the importance of New Pleuron in the history of Aetolian fortification, it does not play a large part in the political history of the nation : it gives only one Strategos to the Leagued The reason of this insignifi- cance is very clear. The site possesses no strategic value ; it lies on no main thoroughfare. Early history had proved Old Aetolia to be a nonentity in politics, but the strategic and economic disadvantages of that site were accentuated in the situation chosen for the new city^. The treaty with which Rome completed the ruin of Aetolia in 189 b. c. gave the town to the Achaians. We learn from Pausanias that Sulpicius Gallus allowed an Aetolian embassy to go to Rome to pray for a dissolution of the odious connexion *. Rome, then no longer apprehensive of Aetolia, granted the boon as a means of breaking in its turn the Achaian power. After this, New Pleuron is heard of no more, except as a name in Aetolian geography^. OLD PLEURON. Looking southwards from the walls of New Pleuron, we see below us two lesser heights separated by a depression from each other, and by a torrent and narrow waste arm ' The Laphrian Artemis is scarcely an exception, as she is found on the coins of Patrai. ^ Namely, Pantaleon ; who was, however, Strategos on three occa- sions, — 186/5 ; 1^/79 (') ; 173/2 B. c. Cf. the inscription from Mokista, on p. 207. ' Hence Philip V, who in 219 b. c. passed by Pleuron in order to ravage the Kalydonia, struck more truly home than did Demetrios in 235 B. c. Was the title AtVmXucdr given to Demetrios only in mockery by those who recognized the futility of his Aetolian exploit ? We see from the considerations above suggested that the gift of Pleuron to the Achaians is only another instance of Roman astuteness. * Paus. vii. II. 3 : 'A(J>Uovto Se as t6v rdWov Koi AlraKSiv ol n\fvpS)va oIkovvtcs, (rvvreXeias Ttjs is 'A)(ai.ovs i0e\ovTes nipfiTiv evpaaBai. koi airols iirtTpawf] n^v ino Tov FaXXou jTpetT^ciav em atpav airav ISia irapa Paiuuovg aTroareiXai, fweTpdiTti 8e iiTO 'Pa/iaiciiv trvvtbplov tov 'AxaiHv artoaTrjvai, ° But Pleuron had the honour of giving birth to the only Aetolian poet known to us,— Alexander, son of Satyros and Stratokleia. His name was included in the group called 'H nXaar. About 285 B.C., in the reign of Ptolemy Philadelphos, he was in the Library at Alexandreia. 128 THE COAST PLAIN. [cH. ^ of the plain from the slopes of the Zyg6s upon which we stand. The nearer hill is called Gyphtokastron, the ' Fortress of the Gipsifes ' ; the lower one beyond it is Petrovuni, ' Peter's Hill '} About half-way up the side of Gyphtokastron an irregular ring-wall, scarcely half a mile in circuit, encloses the entire hill. There is the greatest possible contrast between the style of this enceinte and that of New Pleuron. Thin slabs and irregular blocks with little trace of squaring are loosely piled to make a wall that might as easily be taken for the hasty and careless work of a modern peasant as for the production of a high" antiquity. Here and there it assumes an appearance of greater elaboration and strength. Yet a third style seems to be traceable in certain parts of the enceinte, for conspicuous among the small stones of which it is composed are ' some very large wrought stones, the work apparently of a remote age,' as Leake observed ^. Repairs dating from post-Hellenic times show that the kastro has often been occupied since the days of antiquity, as indeed we should infer from its modern appellation. At present the site serves chiefly as a quarry, and the stones are being steadily carted away to Mesolonghi. Consequently, the defences are nowhere more than two or three feet high; generally their course is only to be made out with difficulty over the rough ground. Within the enclosure there is nothing, save a few inexplicable remains on the top of the hilP. More interesting are the ruins on the smaller , height, that of Petrovuni. From the north-eastern corner of Gyphto- kastron a long wall runs obliquely down the slope towards the hollow between the two hills. In this depression there are traces of a large and important gate. The wall continues its course up the slope of Petrovuni, then bends and runs along its eastern face, just below the brow of the hill, until in the south it reaches ground so rugged and broken that its further extension becomes both impracticable and unneces- sary: in this quarter no enemy could approach over the rocks in force sufficient to imperil the safety of the garrison. Owing ^ Tv(l)T6Kaarpov. Uerpofiovvi. " N. G. iii. 539. ' If the temple of Athena was in the midst of Old Pleuron, these remains might belong to it ; they seem to be those of a building of some size and importance. x] PETROVUNI. 129 to the damage done by the quarrying operations it has become impossible to say whether a connecting wall ran from the south-eastern corner of Gypht6kastron across the hollow in correspondence with that on the north, but it seems likely, as otherwise the connexion actually in existence would fulfil no practical purpose. The blocks in the walls of Petroviini are generally of huge dimensions, and they have been placed in position without much attempt to improve upon the surface produced by the natural lines of cleavage. Smaller stones are used for the purpose of filling up the interstices. The most important feature is the gateway in the road between the hills, but it is too much damaged to enable its exact plan to be recovered. The opening appears to have been about eight feet wide and thirteen feet long : there are no signs of special defences near it. This sim- plicity is on a par with the thoroughly unscientific character of the fortifications on this hill of Petroviini. Reliance has been placed upon the solidity and size of the wall, not, as in a later and more advanced age, upon the application of prin- ciples that made a gigantic scale of work unnecessary. What inferences may be drawn from the appearance of the ruins themselves ? It is clear that, although of different dates, the remains on the two hills belong to a single scheme of fortification. Further, the traces of very early work noticed on the larger hill suggest the suspicion that the ring- wall at present there existing marks a second chapter in the story of the town on the height of Gyphtokastron ; that the course of events has been something more than a mere peaceable extension of the polls from one hill to the other. Some such hypothesis is required to account for the manifest differences of style and the visible advance in the art of fortification : the primitive simplicity and grandeur of the walls of Petroviini were replaced on Gyphtokastron by the complications of science. Though its walls are more frag- mentary than the remains that we have hitherto examined, we seem on Petroviini to trace the dim outlines of some episode in the early history of Old Aetolia, and to stand in the presence of the most distant past of the nation. The question of the identification of these remains is one of extreme difficulty, and it has not yet been satisfactorily handled. The close proximity of the kastro of Irene, the K I30 THE COAST PLAIN. [ch. obvious sequence in date in the masonry and principles of fortification on the two sites, agreeing so well with the known facts of the history of Pleuron, — such considerations suggest that we should see in the remains on Gyphtokastron and Petrovuni the relics of that earlier Pleuron which was captured by Demetrios of Macedon. Such was the opinion of Leake '^, so that our question would appear to have been answered almost before it was asked. The difficulty, how- ever, is that this identification is apparently not in perfect accord with the indications furnished by our authorities! They seem to point to another identification, that adopted by Bazin. He maintains that the ruins are those of Olenos, one of the five Aetolian cities known to Homer ^. The testimony of Strabo with regard to Old Pleuron is contained in three passages. In the first we read that the mountain Kourion was ' near Old Pleuron ; some imagine that the people of Pleuron gained from it the name of Kouretes^.' And we have already heard that Old Pleuron was ' situated in a fertile plain, a short distance from Kaly- don *.' We must be on our guard against unduly pressing these last words in the interests of a particular theory. This has been done by those topographers who put Old Pleuron on the alluvial plain by the sea, as for example on the site now occupied by Mesolonghi ^, or who are inchned to look for it considerably to the east of Gyphtokastron. The first position is quite out of the question, and the second exagger- ates the force of Strabo's words. The Kouretes gave their own name to the mountain, rather than were -called after it, but otherwise there is nothing in these two passages to militate against the identification of Old Pleuron with the ruins at the foot of the Zygos. Confusion and doubt are first introduced by the remark of the geographer concerning the course of the Euenos : — ' The Euenos does not at first flow 1 N. G. i. ii8 ; iii. 539. " Baz. Mem. p. 353. ' Str. p. 451 : Kovpioi' bi ir\rjirlov Trjg naXaias 0\evpS>vos, deft' oS tovs QXevpto- viovs KovprJTas ovofiatTB^vai Tives tureXajSoy. Steph. Byz. makes his usual mistake, of calling Kourion a ttoXii. See also Strabo, p. 465 : 01 S' anb rov opovs Tov KovpLov . . . Tov vTTepKfi/jjevov T^s nXfupwvos. * Str. p. 451. See p. 124. ° So Bazin, Mem. p. 354 : ' Un peu au dela de Missolonghi, par exemple dans la plaine, au pied d'un des contre-forts du Zygos,' which is, how- ever, vague enough to suit any site. x] SITUATION OF OLD PLEURON. 131 through the Kouretic territory, which is identical with that of Pleuron, but more to the east, by Chalkis and Kalydon ; then it turns west, and bends towards the plains of Old Pleuron, finally making a sweep southwards to the sea ^.' With respect to the course of the river, Strabo is indeed partially right : his mistake is that of an early geographer, who has little sense of proportion, and gives to small facts an importance which fuller knowledge denies them. Without specifying the point of the compass from which it flows, he says that the Euenos runs first east, then west, and finally south. This is, in fact,, a fairly accurate description of the direction taken by the Phidharis in its lower -reaches, — of which alone Strabo or his authority would have any know- ledge. From the bridge of Dhemitrakakis at Vlachomandhra, where the Phidharis is crossed by the modern road leading from Naupaktos into Central Aetolia, the course of the river traces a zig-zag figure to the sea. Its direction is at first nearly south ; then follows a bend westwards, with a final turn to the south along the foot of Mount Varassova. The remark that the river approached Chalkis before turning west is suf- ficiently true if we take Strabo to have conceived its course as being more nearly parallel to the coast-line than is actually the case ; a mistake that is, in fact, not impossible even for the modern traveller as he rounds the shoulder of Varassova. There remains the question of the situation assigned to Pleuron and Kalydon in the passage before us. If the Euenos in changing to a westerly direction came into the plains of Pleuron, then there is no alternative, — the plains of Pleuron can be none other than those of the modern Bochori, on the west or right bank of the river ; and Pleuron, consequently, must have possessed all the land upon that side between the sea and the Zygos ^- It is true that when he goes on to speak more minutely Strabo confines himself to saying that the Euenos 'turns in the direction of the * Strabo, p. 451 : pel 8' ov Sia rrjS KovpriTilirjS KOT apxds, ^ris iariv tj avrr) TJj 'n\eupaviq, aXKa Sta rijs Trpoacaas fiaXKov irapa rrjv XaXKiSa (cat KaXvSava' eh' avaKap.'^as iid rh Ttjs n\evpS>i/os ireSia rrjs naXmas Kal irapaWa^as els 8v(rii> eiridTpe^ei itphs ras eK^oKas Koi Trjv fiea-r^fi^piav. 2 And, in spite of its inherent absurdity, this conclusion was accepted by Lolling, who says that Pleuron ' nahm die ganze Kustenebene bis zum Euenos in Anspruch.' K 2 132 THE COAST PLAIN. [ch. plains of Old Pleuron,' but he spoke before of its ' flowing through ' them ; and that is undoubtedly what he meant. The reference to Chalkis and Kalydon was dictated by a desire to guard against the possibility of his readers extending the Pleuronia too far towards the north-east; which, of course, they were liable to do if they were not warned that in this direction room was to be left for the territories of Chalkis and Kalydon. The conclusion is inevitable, therefore, that Strabo assigned to Old Pleuron the plain at the foot of the Zygos on the right bank ofthe Phidharis where it issues from behind Mount Varassova. If so, it is no less inevitable that he here identified Old Pleuron with the ruins existing at that point, those which are undoubtedly to be attributed to Kaly- don. Forced into error by the effort to reconcile conflicting authorities, Strabo has made his geography of Southern Aetolia a chaos. Once having spoken of the Euenos as flowing through the Pleuronia, he was compelled to identify the kastro of Kurtaga with Old Pleuron, and to put Kalydon upon the left bank of the river. The truth is that it flowed through the Kalydonian plain, and had no connexion what- ever with either New or Old Pleuron : and Strabo knows this well enough in other passages. This third passage of Strabo has, therefore, turned out to have no claim to rank as evidence on the question as to the true situation of the earlier city of Pleuron. And the two statements previously cited ^, although true, are of little positive value, seeing that nothing remains to enable us to identify Mount Kourion, or to fix the distance of the old town of Pleuron from Kalydon. If, however, the ruins at Gypht6- kastron and Petrovuni are assigned to Old Pleuron in accordance with the suggestion of probability eked out with those two vague passages, then a new difficulty must be faced, in the claim put forward on behalf of the Homeric Olenos to be located upon the same site. Olenos also is thrice mentioned by Strabo. He says : ' The Aioleis destroyed Olenos, which was near New Pleuron, and the Akarnanians disputed its territory with them^.' ^ On p. 130. "^ Str. p. 451 : Trjv )uv "OXcvov . . . AioXcis KaTe(rKaylfai>, ir\r)a'i.ov ovcrav tSjs vetorepas TlXevpavos, t^s Si }(a>pas ruitpur^rjTOVv 'A.Kapvaves. x] WHERE WAS OLENOS ? 133 Farther on he adds : ' In the interior of Aetolia . . . was Olenos, which is mentioned by Homer in his list of Aetolian cities ; there remain only vestiges of it near Pleuron, at the foot of Arakynthos ^.' His third passage is simply a repe- tition of part of this ^. Leake's view^, that Olenos should be sought on the borders of Lake Trichonis, — that it was, in fact, upon the site afterwards occupied by Trichonion, — is due to his insisting upon the expression ' in the interior.' It has already been noticed* that the cities of Homeric Aetolia lie all appa- rently within the same zone, — the maritime plain on the southern side of the Zygos. There is nothing to lead us to make Olenos an exception, nor is there any evidence that Trichonion was built upon a site that had previously been occupied ^, or that it was included within what Leake- calls ' the dominions of the Kalydonian dynasty.' This last ex- pression seems to have been suggested by the idea that the kingdom of Kalydon was co-extensive with Aetolia 'Apxaia, than which there can be no greater misconception of the early history of the country. ' In the interior ' must only be taken to imply a situation not actually on the coast ^. The fact that the Aioleis, into whose hands Olenos fell, were involved in border wars with the Akarnanians points to a site very near the Acheloos. The conclusion to be drawn from the above-quoted pas- sages admits of no doubt. If Strabo has not gone entirely astray in the data he gives, the claim of Olenos to be identi- fied with the remains below the kastro of Kyria Eirini must ^ Str. p. 460 : Trjs 8e /ieiToyaias • • . Kara tiji' AhaKiav rfv QXevos, • . . "x"'! ^' avTtjs \el7rera1 /wvov iyyiis r^s HKevpSivos vtto tte 3. ^ Just as Pliny {H. N. iv. 3), speaking of Pleuron, without doubt the later city at Kyra-Eirini, says that it is ' in the interior ' :— et in mediter- raneo Plevron Halicyma. The occurrence of this last name enables us to attach its true value to the phrase. 134 ■ THE COAST PLAIN. [ch. be held to be established ^ Yet, if Olenos also must be placed at Gypht6kastron and Petrovuni, what becomes of the old town of Pleuron ? We have reached the following position ; we have a single site, which we have successively identified with Old Pleuron and with Olenos. And both identifications are sound. What is the explanation of the paradox ? It amounts simply to this : that the Olenos known to Strabo was not the Olenos of Homer and early AetoHan history. Strabo and his informants were the victims of a topographical fraud. Olenos, it must be remembered, had been destroyed cen- turies before Strabo wrote. In the chiaroscuro of the epoch of the migrations it is swallowed up by the flood of invading Aioleis. Further, the site of Old Pleuron also had been deserted two hundred years before the time of the geographer ; and if we may judge from the vagueness of his language with regard to Old Pleuron, as contrasted with his distinct and minute assertions respecting Olenos, the former town had vanished from the face of the earth. Yet Hellanikos, in the middle of the fifth century before our era, speaks of Olenos, together with Pylene, as being still in existence. It will require something more than Strabo's criticism to convince us that Hellanikos was in error ^. We must accept it as a fact, that during the fifth century the name of Olenos was once more found upon the map of Aetolia. It cannot, however, have attached to the site at the ' This result seems to suffer shipwreck upon the statement of Strabo (p. 460), that Lysimacheia was near Olenos {^v 8e koI Ava-ifidx^ia irk-qa-iov). Lysimachela being certainly on the northern side of the Zyg6s, in Central Aetolia, no possible geography can bring Lysimacheia and Olenos into juxtaposition. It would seem, then, that we are reduced to a choice between mutually exclusive alternatives. Either Olenos was close to Pleuron (Str. pp. 451, 460), or that connexion must be given up and we must look for Olenos on the shores of Lake Trichonis in the neighbour- hood of Lysimacheia, as Leake imagined. But a solution of the difficulty is not impossible. Lysimacheia is also brought into connexion with Pleuron (Str. p. 460 : fiera^v TtKevpmvos Kai ' Apcrivdijs nSKecai) . It thus becomes easier to understand how Olenos and Lysimacheia should have been mentioned together, seeing that Pleuron also, the near neighbour of Olenos, stood in some geographical relation to the same town. See p. 223. ' Cf. Strabo, p. 451 : 'EWaviKos 8' oiSe r^v rrepi ravras 'urTopiav olSev, dXX' as en KOI avT&v ovtrmv iv lij apxl>v 8e 'AvTifjaxov mreKTfivev &Ka>v 'HpaxX^r olvo^oovvra air^, as i^iKavSpos laropei iv hevrepa Olrdinav, ^ Kol dveicrBai £TU' KoXoviiivrjv X<"P<"') V" ° TOT-Ojuor cTTifcXuffi, ireptfiAxriTov eiroia to iraXoiiv Toiis opovs avy)(jiov(Ta del Toiis cmohaKVVjiivovs rois 'Axapvaai Koi Tois AItioKoIs. See also id. p. 459. * Cf. Heuzey, Le Mont Olympe. p. 423. The district on the Akarnanian side of the river is now called To Karm/^cpoE (the Low Country). LIMITS OF THE PARACHELOITIS. 151 As the Paracheloitis is represented as having been the scene of constant warfare, cdused by the real or imagined fluctuations in the course of the river, it is not likely that the term should be extended so as to include the plain of Stratos, for the course of the Acheloos through the central Aetolian basin is not liable to variation. The river issues from the gorge in the hills hard by Stratos in the north, and flows in a well-defined broad bed towards the most westerly of the only two possible exits through the mountains to the' south. The only alternative to its present course is the opening of the Kleisura into the head of the Aetoliko lagoon. Pouqueville, indeed, speaks of traces of an old bed below Stamna; he seems to mean the passage at the foot of Mount Katsa through the hills that bound the western side of the lagoon^. The configuration of the country is such as to make it impossible to regard that opening as having constituted anything more than a lateral -channel between the early mainland and the island once formed by Katsa and its companion hills to the southward. The changes that led to feud are thus possible only after the river has issued upon the maritime plain, and it is to this part that we must confine the name Paracheloitis. Even within historical times the course of the river in its lower reaches has suffered con- siderable modification, and it once entered the sea much nearer Mesolonghi than is now the case ^- In a district like the Paracheloitis, — of a fertility that made its possession an object of keen conflict, but liable at any moment to be snatched from the victor's grasp, — no great city was likely to arise until the question of ownership had ' Pouq. Voy. iii. 519 : ' On reconnalt encore, lorsqu'on en est prevenu, les traces de ce canal au-dessous de Stamna' {sc. the canal made by Herakles, Stf. p. 459 !). Millingen (Memoirs, p. 47) refers to the same :— ' (The) Aspropotamo is continually filling up some of its outlets into the sea, and has thus gradually changed its original course. One of its old channels is plainly traced from Anatolico Bay to near Catochi ; and the cluster of small islands, on which stands the town of Anatolico, owes no doubt its origin to the Acheloos, from which it is now more than eight miles distant. When the inundations are considerable, the river flows again along its former bed.' The old bed referred to in these passages is now marked by the carriage-road from Aetoliko to Neochori (opposite Katochi), from which a road branches off to the right to Mdstru. ^ Dodwell, Tour. i. 102. 152 THE PARACHELOITIS. ' [ch. been settled. This conclusion is borne out by the facts of the ground. As already observed, we must recognize that, powerful as the Aetolians subsequently became, they had a hard fight with the Akarnanians during the earliest period of their history, and that for long the balance of advantage lay upon the whole with the latter ^ Conse- quently, on the Aetolian side of the river we find only a few strong towns, which are, indeed, little more than forts by means of which the Aetolians kept what uncertain hold they might upon the rich plain. On the Akarnanian side, on the other hand, success gave birth to the city of Oiniadai ^, one of the most important places on this coast. At last the tide of fortune set steadily in favour of the Aetolians. The Paracheloitis was gradually subjugated. Oiniadai, long before she actually severed her connexion with the Akar- nanian League, fell a prey to the Aetolian influences by which she was surrounded^. This at least is the most probable way of accounting for the stubborn resistance offered by this city, first to Perikles, and afterwards to Phormion and his son Asopios, in face of the otherwise unanimous adhesion of the Akarnanians to the cause pf Athens *. In course of time the inclusion of Oiniadai in the Aetolian Federation was open and professed, and this explains the continued absence of any great city on the eastern bank of the Acheloos. Oiniadai served the Aetolians as the capital of the Paracheloitis, just as she had previously served the Akarnanians. ^ See p. 140. ' For a description of Oiniadai, see Heuzey, op. cit. p. 435 fol. ' Corinthian influences may also have been at work, as Curtius surmises in his ' Studien zur Geschichte von Korinth ' in Hermes, vol. x. He is compelled, however, to admit that no Corinthian type appears upon the coinage of Oiniadai {ibid. p. 243). Its coinage is local, though its types subsequently became national. The absence of Aetolian influence upon the coins is due, of course, to the fact that the Aetolian Federal coinage was of comparatively late origin. * To Perikles, Thuc. i. iii ; to Phormion, id. ii. 102 : Ohm&as M wore TToXe/iiovs SvTas fiovovs 'Axapvavav ; to Asopios, id. iii. 7. See also Thuc. iii. 94 ; iv. 77 : Ainioa-demis . . . OlvidSas vw6 re 'AKapvdvav wdvrav (tarij- payKaaiievovs KaraXajSrai' h TfjV 'h6r)vaiav ^vfiftaxiav. In Paus. iv, 25. I the Messenians of NaupaktOS rjma-Tavro ObiiSas '&.Kapvdvaa> y^v re ixovrag ayaaqv KOi Xarivaiois 8iav Si Ta irrfva to \oiir6v ^Srj ^dSrjV xal irpaelav Inoielro rijv TTopeiav, ' Kara Se tjjv tripurcurav KoratrKtvifv oIki£>v Kal r€(;(wv Kal nipyiav oiS' onoias rJTTS). 154 THE PARACHELOITIS. [ch. must have been after the Macedonians had been through the land with ruin and death one and twenty centuries ago. PhiHp, marching through the plain below Stratos, crossed the Aspro at the point at which it begins to flow between the western spurs of the Zygos and the forest-covered heights of Manina^ This hilly strip intervening between the upper and lower Aetolian plains is only about seven miles across, but, partly from its narrowness and partly by reason of the level expanse on either side, it is a well-marked and striking feature. The narrow limits within which the Acheloos is confined after passing through the upper plain fully justify the expression 'the Straits,' which Polybios applies to the belt. The rapid march of the ' king through this scrubby country was a measure of ordinary prudence. The capture of Ithoria released him from all anxiety as to his position. We must look for Ithoria, therefore, at the southern end of the defiles, somewhere in the vicinity of the conspicuous village of Stamna, which stands on the ridge between the Aspro and the head of the lagoon of Anatolikon. The lofty rugged cone rising about two miles to the south of Stamna, above the small village of Haghios Ellas ' at the Almond trees ^,' at once attracts the eye roaming in search of a Hkely site. Here at least one condition, that of natural strength, is clearly fulfilled. The rough country along the Acheloos comes to an end at the foot of this hill, and then the plain begins, gradually broadening out beyond the village of Guria, which lies hard by the left bank of the river, two miles to the south-west of Haghios Elias. The identification does not rest solely upon this consideration. The peak contains many traces of having served as a forti- fied post, in the shape of walls that rank among the oldest specimens of military architecture in Aetolia. The hill of Haghios Elias rises steeply a quarter of an hour south of the village. ' We should most accurately describe it as a ridge, running from north-east to south-west and com- . posed of three members, — a steep central eminence with a lower height projecting from it on each side. Between the central peak and the north-eastern prolongation is the ^ For the ford, see p. 210. 2 "Ayios 'HXt'ar arah MuySaXtaiV. For Stamna, see Leake, N. G. iii. 544. p o in o z; o o tn o z o s >-) < w a H < a o < O xii] THE RUINS OF ITHORIA. 155 gateway, and it is here that the fortification is in the best state of preservation. The wall at this point runs in a straight line to the rocks of the central height, and on the other hand towards the crags that form the extremity of the projec- tion in the north-east. Even this section remains to a height of at most only three courses, of large blocks irregularly piled, exactly in the style of the more ancient portions of the remains at Gyphtokastron and Petrovuni. Where the wall met the natural rock a bed was carefully cut for the' stones, so that along the face of the crags the line of the enceinte can easily be traced. On the sloping rock at the summit of the hill three such cuttings resembling steps are called the ' King's Seat ' by the peasants ; and they form indeed a regal throne so far as prospect is concerned. The great summits of western Akarnania, and the higher peaks on the distant confines of Epiros, make the background of the panorama; at our feet, unfolded like a map, lie the lower hills, the plain of the Paracheloitis, and, threading its way through both, the loam-laden stream ^ of the Acheloos. Turning to the east, we gaze over the lagoon of Aetolikon and the plains that extend southwards to Mesolonghi. The Zygos impedes the view on this side, hiding all the interior of Aetolia ; we look into the green folds opposite and note the contrast between the richness of the Lower and the burnt nakedness of the Upper range, severed from each other by the chasm of the Kleisura which we make out on the far side of the water. The gateway by which we enter the fortress exhibits, at first sight, nothing very distinctive in its construction. It is a simple opening, thr-ee . and a half feet wide, with a passage about ten feet deep, the breadth of the wall itself. On the right, just under the steep rocks of the central height, an irregular bastion-like projection from the face of the wall, or rather from the face of the cliff", serves to flank the gateway, though the saillie is very slight, being about one eighth of the interval between it and the gate. The face of the projection is turned obliquely towards the path by which an enemy would approach the gate ; so that, if we regard the wall and the rock above and beyond it as a single line of defence, the projection turns out to be ^ Thuc. ii. 102 : pevna doXcpoV. 156 ' THE PARACHELOITIS. [ch. a real demi-bastion placed- at the angle made by the lines at this point. The fire of the defenders stationed on the projection was thus enabled to flank both the wall in which the gateway is pierced and the base of the rock, — the former in a very effective manner owing to the fall of the ground from the bastion to the gate : at the same time they them- selves were covered by their friends on the rock above them. We now see some reason both for the slightness of the projection and for its distance from the gate, the weak spot in the line. The first is due to the steepness of the slope downwards and outwards, and the desire to keep well inside the range of the missiles of the defenders on the main rock ; the second to the desire to take advantage of the angle for defensive purposes. It is true that the gateway might have bpen placed a little nearer the bastion ; but the nature of the ground makes its actual position the best that could be chosen. Nevertheless, the width of the interval between the demi-bastion and the entrance did not fail to strike the engineers of the place. This is proved by the existence of a short inner line of wall covering the entrance on the inside, exactly in the manner of the modern 'traverse.' For just within the gate the rocks rise in a slight ridge, which has been strengthened with masonry, thus creating a passage about twenty feet wide between the inner and the outer wall. The main wall follows the outer face of the ridge towards the right and left hand from the gate, but the crags of the central hill and of the extremities of the ridge rendered artificial defences needless at those points. The three members of the hill are really disposed on the arc of a circle, with the concavity turned towards the east and the lagoon. The whole ridge is exceedingly narrow, and sinks towards the lagoon in a steep swell, on which the ancient town was built, enclosed by a wall running in a circular sweep from each extremity of the akropolis. At the head of the slope on which the houses stood there seems to have been a cross- wall severing them from the citadel, so that the latter was nothing more than the rugged crest of the hill. The wall of the town itself is scarcely traceable ; in the south-west we find the fragments of a gateway in the same massive style as that of the akropolis. xii] THE HOMERIC OLENOS. 157 The only object of interest still to be described^ is the cistern mentioned by Leake. It is found at the head of the before-mentioned slope, just at the eastern foot of the crags constituting the summit of the hill. Leake gives the vertical section^, but not quite accurately. The mouth, cut in the rock, is four feet nine inches square ; the present depth of the excavation is over sixteen feet, but it is much choked with rubbish. A fine good stucco, of small round stones mixed with cement, lines the interior ^. It is now time to redeem the promise made in a previous Chapter with respect to Olenos*. The conclusion there reached was that Gyphtokastron and Petroviini mark the site of Old Pleuron. After its abandonment in the third century b. c, the ruins of Old Pleuron came to be looked upon as those of Olenos, and Olenos was the name actually borne by the restored town that existed for a time on the two '' The inscribed stele once lying in the pavement of the church of H. Ellas has recently been destroyed, during the restoration of the building. Bazin saw it. It read : EYOYAAME XAIPE. See Mint. App. No. 12. ^ N. G. iii. 552. ' From Leake's description of the ' extensive and interesting prospect ' visible from the hill (N. G. iii. 552), we naturally suppose that he had ascended the peak. Nevertheless, he is so far from suspecting this to be an ancient site that he identifies Ithoria with another Haghios Elias, to the north of Stamna (N. G. iii. 577). That H. Elias is ' nearly opposite to the ruined town at Palea Mani,' and Leake says that he heard of Hellenic remains there. A comparison with two other passages makes it quite clear that the H. Ellas which Leake identified with Ithoria is quite different ' from the k^stro we have described. After describing Palais M^ni, he says (iii. 528) : ' On the opposite side of the river stands a small tjiftllk and pyrgo called St. Elias, around which the lower falls of Zyg6s reach to the river side.' The two places are distinguished in iii. 552, — ' St. Elias, two miles to the southward of Stamna, is distinguished from the tjiftlik of the same name on the left bank of the Aspro, opposite to PaleS Mani, by the name of St. Elias at the Almond-trees.' It is clear that Leake did not personally examine either place, and that he transferred to the H. Elias lying three or four miles north of Stamna, and three miles south-west of Anghelokastron, what he was told of the H. Ellas at the Almonds, viz. that it contained Hellenic remains. The name Haghios Elias is applied to almost every conspicuous height in Greece, so that confusion is easy. We see now how Leake's section of the H. Ellas cistern came to be inaccurate. There are no Hellenic remains at the northern H. Elias. * See p. 136. 158 THE PARACHELOITIS. [cH. hills ^. It was this spurious Olenos that deceived Strabo. He was ignorant of the fact (known to Hellanikos) that the genuine Homeric Olenos had itself also been raised from the ruins in which it had been laid by the Aioleis ; much less did he dream that its name, dissevered from its true home, had seized upon the site that properly belonged to Old Pleuron. We now advance the conjecture that the genuine site of Olenos is the hill of Haghios Ellas. The town known to Polybios as Ithoria was the lineal descendant of the Homeric city. True, we have not a scrap of positive evidence to adduce in support of this hypothesis, but it is not entirely visionary. The ruins themselves give us tangible facts. On the one hand, the fortifications on Gyphtokastron belong to at least two distinct epochs; on thd other, the masonry at Haghios Elias at the Almonds indubitably belongs to the earlier of the two, exhibiting all the characteristics of that primitive work which in Aetolia is confined to three or four examples. If Olenos stood at Haghios Elias, we can under- stand why the Akarnanians should have seized the opportunity afforded by the Aioliart invasion to swoop down upon its territory I We can also see how the city could be described as being ' in the interior of Aetolia ^.' In his compilation Strabo has combined a reference to the genuine with a reference to the spurious Olenos *- It is generally supposed that the invasion of Philip is not the first occasion on which the name of Ithoria appears in history. Diodoros tells us how, in 314 b. c, Kassander ' We have, of course, no literary evidence of the restoration of the town on the hills of Gyphtokastron and Petrovuni ; it is known only from the remains themselves. ^ Str. p. 451. For the fruitfulness of its territory, cf. Stat. Theb. vi. 423 : nee Oleniis manant tot cornibus imbres. Here there is evidently an allusion to the position of the town in the rich plain reclaimed from the Acheloos, represented in the legend as a horn broken off in his struggle with Herakles (Strabo, p. 458 fol.). Statins may be reproducing some proverbial expression which has not descended to us. ' Strabo, p. 460. * The reference to the genuine Homeric Olenos on the hill of H. Elias is contained in the words t^s be iieaoyatas . . . Kara t^v AtTcaXiav ^v "OXevor (p. 460), and T^s 8c x^P"^ rjiupur^rfrow 'AKapvmes (p. 451). That to the spurious Olenos, on the hills Gyphtokastron and Petrovflni, lies in the statement nXrja-iop oSa-av T^s veayrepas WKeupavos (p. 451), which is repeated in the eyyis Trjg HXfvpavos iiro Tc5 'ApoKvvBa of p. 460. xii] OLENOS AND ITHORIA. 159 marched into Aetolia to the support of the Akarnanians^ Following his advice, the Akarnanians concentrated their forces by withdrawing from the smaller towns. Most of them gathered to Stratos ; the people of Oiniadai, along with others, to Sauria ; the Derieis to Agrinion. The '2,avpiav of the Greek text is usually emended to 'WpCav^. It .is impossible to decide the point; but when we recall the desperate and successful efforts of the Aetolians to recover Agrinion we may ask how it is that Ithoria, which also lies on the left bank of the Acheloos, and, therefore, geographi- cally belongs to AetoUa, is not heard of as sharing the fate of the more northerly town. THE FORTS. Of the small forts mentioned by Polybios in connexion with the raid into the Paracheloftis and meagrely described by Bazin ^, Leake apparently had no knowledge. The dis- covery of their remains furnishes a valuable proof of the minuteness and accuracy of the information upon which Polybios relied in questions of Aetolian topography*. The ruins of the first of these forts are hidden among the brambles and thick undergrowth in the woods on the left bank of the Aspro, two or three miles north of Stamna. A narrow strip of flat land runs between the river and the hillock on which the fort was built ; a short distance farther north are the vestiges of the 'Two Churches,' which give their name to the site. The hillock contains the insignificant remains of a small quadrangular redoubt, the wall of which stands at one corner to a height of three courses, composed of large blocks laid regularly, but with joints oblique. At a distance of a few paces we see the foundations of small buildings among the bushes, but it is impossible to make out their meaning. ' Diod. xix. 67. See p. 172. ^ So Leake, N. G. i. 156, note i. Bursian, Geogr. i. 120, accepts the emendation, but puts both Ithoria and Paianion on the right bank of the Acheloos, identifying the former with the kastro of Palaiomdni ; cf. id. Rk. Mus. xvi. p. 440. Heuzey, L'Acamanie. p. 434, keeps 'Savpiav (= Lizard town) as a town of Akarnania, and identical with Palaiomdni. Lolling also puts Ithoria at Palaiomani. ' Mem. p. 340. * Cf. p. 258 for a conjecture as to the ultimate authority for all that Polybios tells us Iwith regard to the topography of Aetolia. i6o THE PARACHELOITIS. [ch. If we descend from the village of Haghios Elias at the Almonds in a direct line westwards to the Acheloos, we reach the Helleniko, as the remains of the second fort are called. They also lie on a slight elevation, the last towards the river, which flows at only a few yards' distance. Only a few stones of the two lowest courses are left in situ, but the outline of the small enceinte is easily traced by means of the carefully cut beds prepared for the blocks where the rocks project above the surface of the hill. Its plan is that of a quadrangle, about thirty-eight feet long and twenty-seven feet wide, with its longest axis running east and west. As in the first fort, in the thickness of the wall there is only a single stone, measuring more than two feet across and two feet in depth ; the foundation course is only half that depth, projecting slightly beyond the face of the wall, as is often the case in Hellenic masonry of the best kind. The third redoubt is about half an hour's walk due south down the river in the. direction of the village of Guria, which stands conspicuously on a low height close to the Aspro. A similar but much smaller knoll a hundred yards from the river bears the fragments of the fort, in the midst of which are the ruined walls of an ancient church ^. Haghios Ehas is half an hour away to the north-east ; to the south, a dusty plain of soft yellowish clayey soil extends to the hill of Guria ten minutes distant, and continues beyond it in the direction of Mastru. So far as can be made out, this fort also was a quadrangular enclosure lying east and west, but it is only at the north-western and south-western corners that any of the original work remains : the structure has been ruined to provide material for the church. The wall of this redoubt may have been of somewhat greater thickness than that of the two lying farther to the north. The enclosure measured about sixty-five feet in length and fifty-two in breadth. In style of work it precisely resembles the two redoubts first described. These are without doubt the forts in the vicinity of Ithoria that were dismantled by Philip's troops. The word vipyoi, ' towers,' applied to them by Polybios, exactly expresses their character. Their small size excites surprise, and it is im- possible that they should have been seriously intended for ' Of the Holy Apostles, or of St. John. I heard both names. o in O 3 o o o o hi K u < w H a o O H « O XII ] P AMNION. i6i military purposes. Their connexion, again, is with the rivet rather than with the road, and it is evident from the nature of the ground that not one of the three redoubts can have been designed to prevent the passage of an enemy along the left bank of the Aspro. The grandiloquent description given by Bazin ^, that they ' disputed the passage at the point at which the defile ceases and the plain begins,' is neither true of their actual position nor has it here any real significance. They were probably intended to serve as look-out posts over the fords of the Acheloos, and as temporary places of refuge for the labourers in the plain in case the Akarnanians succeeded in stealing upon them. The total absence of fallen or scattered blocks suggests that the original height of the enclosures was insignificant. They were little more than ordinary breastworks, — of course made in stone; for the Greeks, as Polybios points out ^, never developed the tactics of the spade. This conclusion is confirmed by the absence of any trace of doorways or other means of entrance. In the case of such buildings, little more than sentry boxes, there was no difficulty in springing over the parapet, while to prevent an enemy doing the same was an easier task than it would have been to hold a doorway, which, not being flanked in any way, and in so confined a space, would have depended for defence upon the arm of a single man. PAIANION. About one hour to the south of Guria, and the same distance west of Anatolikon, lies the village of Mastru ^- Between the Acheloos and the hill on which the village stands, about a quarter of a mile from either, there rises a rocky eminence of moderate elevation, crowned with remains. Their existence was reported by Leake*. The summit of the hill is level, except for a slight swell in the north-west. The wall runs along the crest, following the configuration of the hill, so that the ground-plan of the fort is ' Mini. I. c. ' disputaient le passage au moment ou le defile cesse et ou la plaine commence.' '^ Pol. vi. 42. ' Mao-Tpou. * N. G. iii. 553 : ' On a projecting point of the Stamn^ ridge, half-way between Mastii and the Aspro, are the foundations of a fortified Ko)/i7, nearly of the same size as those at Skortiis and Prodhromo.' The latter are kSstra of Akarnania. M i62 THE PARACHELOITIS. [ch. that of an irregular polygon, with a circuit of less than half a mile. Unfortunately the enceinte is ruined to the founda- tions; only at a few points can it be found existing to a height of one or two courses. Its breadth varies from eight to ten feet. At- regular intervals of fifty feet, square towers are placed, of which the face measurement varies from twenty to twenty-six feet; they project some twenty feet from the wall. ■ The style is upon the whole a careful form of ' regular Hellenic,' but here and there we find obhque joints and courses cutting one into the other. The hill is not naturally strong, but this has been compensated by the multiplicity of towers, and by solidity and care in construction. Within the enclosure we trace foundations of public or private buildings of the same careful workmanship. Near them, but scarcely distinguishable, are vestiges of a gateway in the eastern wall. Scattered over the site are fragments of tiles. It is scarcely necessary to call attention to the numerous points of agreement between the features of these remains and the description of Paianion in the pages of Polybios. We can only wonder on what grounds Paianion has been sought elsewhere ; as by Becker at the village of Katochi, on the western bank of the Aspro^. He imagines Philip to have crossed the river by the fords of Guria. Heuzey, however, does not mention Katochi as an ancient site. The ruins there seen are Byzantine ; the ancient Greek stele found in the church cannot prove that a town existed on the spot in classical times ^- Nevertheless, Becker , has come ' Diss. ii. 20 : ' quare eodem loco hoc puto fuisse, quo nunc Katochi est pagus, IV mill. gr. ab illorum {sc. Oiniadai) moenibus distans, ubi insignia quaedam, quamvis pauca, antiquitatis monumenta manent.' He seems to have been influenced by the words 5" « Trpbs tovs Olviabas. KaTav ntpi t6v 'Ap^fXaoj/ vrjaiSav t6 ovto ttclBos a(r\ itaBeiv tn t^s vtto tov Trora/ioC Trpajfixreas tov itfKayovs, Kai hlraiKiKai axpai elm, vrjtri^ov(rai Trporepov' K.T.X. The WOrds avyxovvrai Se Koi ai Xoitrai, ms 'HpdSordf i^ijert must be put in parentheses, or omitted. They have the air of an annotation. CHAPTER XIII. Central Aetolia. sites between the aspro and the eremitsas. Advancing from the spurs of the 2yg6s up the Acheloos we do not meet with remains on either bank of the river until we reach the hills on the northern edge of the central plain. There we find the ruins of two towns that were of. great importance in Aetolian history. They are those of Agrinion on the east, and of Stratos on the west, of the Acheloos. Geographically, Stratos of course belonged to Akarnania, but, as also was the case with Oiniadai, its history is bound up with that of Aetolia. For a description and plan of its site reference must be made to the work of M. Heuzey\ The remains on the Aetolian side of the river are known as the Kastro of Spolaita ^, so called from the nearest village, which lies about two miles to the north of the site, among the hills that run back to the river Zervas. The last southern spur of these hills bears the ruins which we identify as those of the ancient Agrinion^. The akropolis height is a semi-isolated ridge, running north and south. The summit of the ridge forms a level platform, with a knoll at its northern end, from which it falls steeply to the bed of the Platanorema*- On the west it sinks no less abruptly ' Le Mont Olympe et FAcarnanie, p. 331 fol. This should be sup- plemented by reference to the account of the French excavations on the site. '■^ 'S.TToXaira. ' Following Bazin, Mem. p. 315. * U\aTav6ppevfui. It gets its name from its numerous plane-trees. It t:ontains practically no water in summer. 170 CENTRAL AETOLlA. [ch. to the plain, which, dotted with trees, extends as far as the Acheloos, about a mile distant. An irregular pentagonal enclosure occupies the whole summit of the ridge, but the area embraced is not large. The course of the wall can be traced throughout its extent. For the most part it is destroyed to the foundation, especially' on the west and north. On these two sides it followed the edge of the steep crags, and consequently did not require any flanking defence. On the south and east, where the slope is more gentle, square towers were employed, — two on the short southern wall crossing the ridge, and six along the eastern line of the enclosure. It is only along the latter side that the wall remains to any considerable height ; in one place we find nine courses, which means about twelve feet^. Its breadth averages ten feet. The style is a careful form of ' irregular Hellenic,' — courses of nearly uniform depth, rarely cutting into- each other, and the joints generally upright. The material employed is the soft brown local sandstone, which has a natural tendency to spUt into quadrangular blocks. Two gateways may be traced, in the eastern wall. One of them is defended by two towers. The other, at the north- eastern angle of the enceinte, opens into a narrow passage, one side of which is bounded by the eastern wall itself. On the eminence at the northern end of the enclosure, and on the platform below it, are seen numerous traces of buildings and terraces similar to those at the kastro of Mastru, to which fortress this of Spola'ita bears in other respects a close resemblance. One of the most striking points about the site is the view obtained from it towards the south and west. The prospect embraces the expanse on both sides of the Acheloos meandering in numerous channels over the broad stony bed which blushes with oleanders or glistens dazzlingly white between Aetolia and Akarnania''. In the ^ The topmost course is about level with the ground inside the enclosure, so that the wall must have been originally at least six courses higher than it now stands. ^ This may have been the origin of the modern name. Aspropotamo = White River. Cf. Leake, N. G. iii. 513. Or the name may have been derived from the colour of the water. Cf. Mure, Journal, i. 102 : ' Its waters are of a whitish yellow or cream colour, similar to those of the < H z H U Z See p. 87. ' ,N. G. i. 155. ' Paus. ii. 25. 2 ; Str. p. 462. * See p. 86. " Diod. xix. 67. 3. ° Diod. /. C. : hieKB^v &n ttoXe/xoi; t^ovaiv ofiopov ix. irdKaxav xpovav. ' 2iii;fj3oij\cueK ck tS>v ava>\vp(av [MSS. op^vpSv] (tai fUKpav xapiav eh 6\iyas TToXeis f»€Tot/c^(rai. ' Oi TrXeioToi fiev els Srparoi' ttoXik trvvt^Kricrav, op^upeoTaxiji; oSaav leal fie^'uTTrfv, Olviabtu 8e KOI nves oXXoi ovvriKBov eVi SaupiW*, Aepieis be pxff ereprnv els 'Kypiviov. For Savpia see p. 159. ' See pp. 140, 152. xiii] THE K ASTRO OF MAVROVRU. iT^ border fortresses was the last desperate resolve of a people conscious that dominion was slipping from their grasp. Agrinion and Stratos give each other mutual support, and command the fords of the Acheloos north of the Zygos, as does Ithoria those south of the range. Of the two positions on the Aetolian side of the river, Agrinion is by farthe most • important. Ithoria, though of surpassing strength in itself, has not to the same extent as Agrinion that desirable immediate command of the plain for the possession and defence of which it was primarily fortified. Agrinion, again, being the only carefully fortified city in the western portion of the central Aetohan plain ^, situated on its very threshold, was invaluable as a base for Akarnanian invaders, but terribly menacing to an Aetolian army attempting reprisals ^- Kassander judged aright when he selected Agrinion for one of the great fortresses, but he defeated his own end. He made it patent to the Aetolians that the duel between them- selves and the Akarnanians must be fought out under its walls. The grim answer to the challenge issued from the king's tent on the banks of the Kampylos was the immediate investment of the town, and the massacre of its garrison as it marched out after capitulation. We have already put forth the conjecture that the hill country north of Agrinion was occupied by the tribe of the Thestieis ^. The two sections of that tribe, the Eiteaioi and the Eoitanes, inhabited the entire tract included between the Aspro and Mount Viena, and between the central plain and the river Zervas. Possibly, indeed, their southern boundary was coincident with the northern shore of the lakes. With equal probability we may take the river Eremitsas to have formed the line of division between the two sections of the Thestieis. The chief town of the section living on the west of the ' Konope (Anghelokastron) was a mere (ca/iij until about 285 b. c. See p. 215. ■■' This might have been strikingly exemplified in the winter of 1822, when the Turks were in full retreat from Mesoldnghi. But the disunion of the Greek leaders caused them to abandon this first line of defence for one more to the west, the Pass of Machal^s. Trik. ii. 378. This pass can of course be turned, as it was in 426 b. c. (Thuc. iii. 106). ' See p. 86. 174 CENTRAL AETOLIA. [ch. Eremitsas must have been the strong fortress now known as the Kastro of Mavrovru ^- The miserable hamlet of Mavrovru lies on the banks of the Zervas, some six hours north of the modern town of Agrinion (Vrachori). On the left bank of the river there rises a steep pyramid, connected towards the south-east with two smaller conical heights. On the south and west the main eminence falls in perpendicular cliffs ; in the course of ages the earth has been washed from the summit above the crags, and is now gathered at their base in a shelving bank overgrown with brushwood and littered with huge fragments of con- glomerate split from the precipices. On the north the hill falls somewhat less steeply towards the Zervas. On this side, just under the summit, there is a small level space, once used as a threshing-floor. The peasants call it rj Uopra, ' the Gate ' : here in fact the entrance to the akropolis must have been placed. The line of the fortifications can be traced from the base of the precipices at the north-western corner of the hill round the edge of the level to the crags of the southern side. On the south and west the sheer rock rendered artificial defences unnecessary. It is impossible to make out the plan of the gateway leading to the platform, for, in addition to the havoc wrought during the building of the threshing-floor, we trace repairs and alterations dating from the Middle Ages. Many fragments of an architectural character are scattered about ^ or used in the repairs, such as pieces of small Doric columns, and slabs with sinkings for I 1 cramps. The style of work is fair ' irregular Hellenic' The material employed is noteworthy. It is throughout a grey sandstone of firm and close texture, whereas the natural rock of the hill is a very coarse conglomerate^- The summit of the akropolis contains no vestiges of antiquity ; like many another Aetolian citadel, it is a wilderness of bare crags and tangled brushwood. The ancient town must have stood on ' Maupo/SpoB = ' Black Water.' Bazin (Mem. p. 316) calls the ruins those of ' Thestia,' the capital of the Thestieis. But he makes no attempt to reconcile the position assigned to the Thestieis with that assigned to the two tribes mentioned on the boundary stone {ibid. p. 317). '' Some of them are ornamented with crosses. These are probably the reported ypafinara for which I searched in vain. ' Geologically the hill appears to belong to the same formation as that of Anghel6kastron. 12 24 36 FEET. B 6. TEMPLE ON ziRVAS. A. PLAN. B. SECTION OF SIDE-WALL. Jo face p. 175. xiii] TEMPLE ON THE ZERVAS. 175 the northern slope running down to the Zervas; a few fragments of the lower enceinte may be traced in this direction. Although of moderate elevation, the hill of Mavrovni, partly from its comparative isolation, partly from its striking outline, dominates the surrounding country. It is in this respect the complement of the still more striking and conspicuous hill of Vlochos, which is visible far down in the south-east on the banks of the Eremitsas. On the banks of the Zervas, some distance west of the kastro of Mavrovni, we find an interesting ruin, the more interesting for the rarity of similar remains in the land of the Aetolians. The Zervas at the point indicated is spanned by a modern stone bridge on the line of a proposed high road between Agrinion and Haghios Vlasis. The road has never been completed, and the bridge is. now a ruin. It is known as Frankoskala, the ' Bridge of the Franks.' It leads directly to a level piece of ground stretching for a short distance along the left bank of the river. On this small plain the remains in question are to be seen. We discover a quadrangular enclosure, ninety-two feet long and forty-three feet wide. At fourteen feet from either end, cross-walls divide the enclosure into three chambers. In one of the partition walls we find a doorway, seven feet wide. The outer wall stands to a uniform height of a little more than two and a half feet. It is constructed of upright slabs, uniform in size, standing on a projecting cillcourse of squared stones ; there are two slabs in the thickness of the wall, which measures two feet in breadth. The longest axis of the building lies from north-east to south-west. The soil brought down from the neighbouring hi'gh ground has almost completely buried the two adjacent sides nearest the hill, and the accu- mulation within the enclosure is on a level with the top of the two sides that are exposed. > Without doubt we have before us a small temple belonging to the Thestieis. There are no vestiges of columns or other architectural members to be seen in the vicinity, but iounda- tions of rectangular buildings are visible a few yards to the north of the temple. The clearance of this small site might yield interesting results, especially as no purely Aetolian temple, or building of any kind, has as yet been excavated. 176 CENTRAL AETOLIA. [ch. The river Zervas ^ is identified by some topographers with the ancient Petitaros^ and their error is perpetuated in modern maps. The Petitaros is mentioned only once by the ancient authorities,— in the passage of Livy wherein are described the operations undertaken by Perseus of Macedon in the winter of 170 b.c.^ When the king found that his design of occupying Stratos had been forestalled, he ' crossed the river Petitaros, and encamped five miles from the city.' We notice that Livy does not say that the Petitaros itself was five miles from Stratos, but that the camp of Perseus, after he had crossed the stream, was at that distance *- We have already pointed out the impossibility of reconciling Livy's description with the supposition that the Zervas is the Petitaros ^ We need only ask how Perseus, being practically in retreat, could have ventured to cross the Acheloos under the eyes of the enemy concentrated in Stratos ^. There is in fact but one stream to the north of Stratos fulfilling the conditions of the narrative. That is the small stream of Kriekiiki, flowing in a south-easterly direction from Vare- tadha, and falling into the Aspro at a distance of seven kilometres'' north of Surovigli (Stratos). The identification of the Petitaros with the Kriekuki river was proposed by Leake. The preference of modern topo- graphers for the identification suggested by Heuzey and ' o Zip^as. ^ Heuzey, L'Acamanie. p. 345 ; Bazin, Mem. p. 295. ' Livy, xliii. 21 fol. * Quinque millia passuum ab urbe trans Petitarum amnem posuit castra. '■' See p. 80. " There are, however, no fords, at any rate in winter, between the bridge on the line of the Vrachori-Karvassards road and the Zervas. We might ask further what was gained by crossing the Acheloos at a point that put the Zervas in his rear. That such must have been the case (if the Z6rvas be the Petitaros) is clear from Livy's statement that the king crossed the Petitaros before pitching his camp. The possible rejoinder, that he crossed it in order to come southwards into Aetolia, is disproved by the fact that the next move was into Aperantia, which district all topographers, including Bazin {Me'm. p. 295), agree in locating north of the Zervas. ' Polybios wrote 40 stades (Livy is of course copying him). This is almost exactly the distance given by Livy (5 R. miles), if we may take a Roman mile to contain, roundly, 1500 metres. xni] IDENTIFICATION OF THE ZERVAS. 177 Bazin seems to be partly due to a misunderstanding of Leake's statement of his own case. He writes as follows^ :— 'At less than two hours above Surovigli (Stratos), the river (Aspro) is joined, on the same side, by a tributary which originates to the eastward of Mount Makrinoro, and at an equal distance beyond the river are the ruins of another Hellenic city, at a village near the right bank, named Preventza. The river I take to be the Petitarus, if this name be correct in the text of Livy, and the ruins those of the town of Aperantia, of which Preventza may be a corruption. Livy indeed seems to allude to Aperantia only as a district; but Stephanus, in reference to the corre- sponding passage of Polybius, which is lost, shows the city also to have been named 'Airfpaz/reta.' It is generally taken for granted that the un-named river to which Leake is here proposing to give the name Petitaros is the large stream which, in modern accounts, appears variously as the Bjakos, Patiopulos, Chalkiopulos, or Syn- dekno. This river falls into the Aspro fourteen kilometres, i. e. eighty stades, north of Stratos ^. It cannot possiblyj there- fore, be the Petitaros. Nor was this the river meant by Leake, for his river is ' less than two hours ' above Stratos. Again, Preventza lies hard by the right bank of the Bjakos ; but, according to Leake, Preventza, and therefore the Bjakos also, is the same distance north of the stream which he calls Petitaros, as the stream itself is north of Stratos. The Kriekiiki stream occupies precisely this intermediate posi- tion ^. And Leake's map proves that he was in no danger of confusing the river of Kriekiiki with the Syndekno (as he invariably calls the Bjakos)*. His remark that Preventza ^ N. G. i. 141. ^ More than aj hours, according to the usual estimate of 30 stades to the hour. ' Bursian, Geogr. i. 140, identifies the Petitaros with the modern Bjakos, and yet {ibid, note 2) says that Leake's river is at least two and a half hours above Stratos. Bursian is hopelessly confused about the two streams, and has transposed the names. On the other hand Bazin (Mem. p. 295) simply misinterprets Leake. But this setting up a man of straw and knocking him down is scarcely sufficient proof that the Petitaros must be sought east of the Acheloos. * His only error lay in putting the confluence of the Syndekno with the Aspro too far to the north. He was therefore obliged to refrain from indicating the exact position of Preventza, which he was aware lay N 178 CENTRAL AETOLIA. [cH. lies 'beyond the river' shows that he is speaking of the Kriekuki river. When he further says that Preventza is ' near the right bank/ he means the right bank of the Aspro, which expresses correctly the position of the village. Thus the Petitaros must be added to that long list of ancient names which the great traveller wedded each to its mate of river, mountain, and town, throughout Greece. The Zervas was known to the ancient Aetolians as the Thestios \ On the top of the ridge overhanging the plain between the ancient and the modern Agrinion we find the remains of a fine round tower. It is now called Palaiopyrgos ^. Only part of the tower still stands, the whole of the side towards the plain having been destroyed. It has an internal diameter of twenty- two feet. The stones are two and a half feet across, and two feet in depth, with a length varying from three to five feet. Naturally there is only one block in the thickness of the wall. Each stone is shaped to the curve. As the tower is now only four courses high, that is six or seven feet, it is impossible to say whether it was roofed or not. No tile-fragments, however, are visible; nor does the number of fallen blocks indicate that the building has ever been much higher than it stands to-day ^. farther south than the place to which, for certain reasons, he assigned the Syndekno. What were those reasons? I think that Leake was confused over the application of the name Tripotamo (Three rivers). In N. G. iv. 253, he says : ' the M^gdhova at no great distance below that junction {sc. with the Agrapha river) falls into the Aspro, at a spot to which the union of a third stream from the mountain of Syndekno to the westward gives the name of Tripotamo.' But the name really indicates the confluence of the combined Agrapha and Agalianos rivers with the Aspro. The Syndekno is much farther south, and has nothing to do with the name Tripotamo. As a consequence of his erroneous sup- position that it formed one of the trio of rivers Leake was led to put the Syndekno north of Kremast^ (cf. ibid.), whereas it falls into the Aspro south of that point. ' This I surmise from the statement, in Plut. De Fluv. xxii. i, that the Acheloos once bore the name Thestios. See p. 86, note 4. In early times the Zervas may have been regarded as the main stream of the Acheloos ; just as, according to Bursian {Geogr. i. 12, note i), the Megdhova — Agalianos was so regarded by Hekataios {Frg. 70-72), and by Sophokles (Strabo, p. 271). ' HaXatdn-upyoi (Old Tower). It is one hour from VrachSri, near the church of Haghia Panaghia Vlach^rina. ' But I have not much confidence in this conclusion in this particular s s o a H o xni] PALAIOPYRGOS. 179 In general appearance Palaiopyrgos reminds us strongly of the quadrangular redoubts of the Parachelo'itis. We remark the same smallness of dimensions, the same striking dispro- portion between the massiveness of the work and the limited area enclosed. The style of the masonry is identical with that of those forts; and here, as there, we see reason for suspecting that the structure had no great elevation. We can have rto doubt as to the object of its builders when we look down upon the great plain cleft from north to south by the 'White River.' From ancient Agrinion and Stratos in the north-west, to Vlochos almost due east, the eye travels through an angle of more than two hundred degrees, while in the pure atmosphere every object in the vast expanse is marked with a clearness that shames the map in our hand. The ancient tower in which we stand to gaze upon this glorious panorama was designed as a post from which might be signalled the approach of an enemy from Akarnania. The isolation of the tower, far removed as it is from any fortress, together with the fact that it is too small to have contained more than the merest handful of men, suggest a further question. To what town was this look-out station attached ? It is true that its splendid situation fits it to be a national sentry-box, and that its signal of danger would put all Central Aetolia upon the alert. For example, the evacuation of the towns that fell into Philip's hands in his advance upon Thermon was possibly directly due to the warning beacon of Palaiopyrgos. Nevertheless, it becomes clear that the tower was erected for some more special purpose when we reflect that the akropolis heights of almost every city on the edge of the plain are only very slightly inferior to Palaiopyrgos as look-out stations. This is especially true of those most in danger from Akarnanian inroads, namely Konope and Lysimacheia. Yet there must have been some town especially interested in and dependent upon the information signalled from the tower. That town was the kastro of Mavrovru in the wilderness of hills north-east of Palaiopyrgos. It was quite impossible from that distant post to keep watch over the most remote instance, as many blocks may have been broken up to provide material for the church. It is obvious that this objection cannot apply to the case on p. 192. N 2 i8o CENTRAL AETOLIA. [ch. and at the same time most valuable part of the territory belonging to it; but, while the crops were growing and until they were safely housed, the sentinels stationed in the tower could swiftly signal to their townsmen the inpouring of Akarnanian marauders. The actual existence of the watch-tower, and our surmise that the domain of the Thestieis extended even into the central AetoUan basin ^, thus explain and support each other. The ancient boundary stone on the banks of the upper Eremitsas has already been mentioned, and its information has been used in the discussion on the AetoUan tribes ^ The point at which the stone is found lies to the south- east of the kastro of Mavrovru. At the village of Skutera, or New Sykia ^, where are some unimportant vestiges of antiquity, the Eremitsas issues from a gorge. The path from Sykia to Barlikeika* and the villages at the base of Plokopari goes over the cliffs along the left bank of the river. About one hour from Sykia we reach a blackened rock called rpjjos 6 (ftovpvos, the ' Old Woman's Oven.' Here, by the side of the track, lies a large natural block, about five feet square. The flat upper surface bears an inscription in deeply cut letters, five inches high, in three lines. Below this a few letters, only half that height, can still be deci- phered. The whole of the lower part of the stone was originally inscribed, but the storms of twenty centuries have effaced the smaller and more shallow letters. A slight fracture has destroyed the final letter of the third hne, but otherwise the stone is perfect. The inscription reads thus :— Eireaimv 'Eoirdvcolv. The legend that gives its name to the locality is perhaps worth the telling. The story of the peasants is of a certain ' See pp. 87, 173 ; cf. p. 57, note i. 2 See p. 87. ' "SKovTspd. The remains to which I refer lie a quarter of an hour from the village, on the right bank of the river, just at the mouth of the gorge. * Mirap\tK€iKa. The ancient path must have been identical with that in use to-day. This explains the site chosen for the boundary record. < H ~S M .PJ H w" o H in >< B! J 8e ejrl woXios ev ^larvoi N iKo\fv[a>']vos NiKLd[Sa\ Uifcovos ^la-rvStv, 6eoKo\evov- aas 'A\Kripoi>v Ayrjra BovKarieijs. Mdprupoi KparidSas 'T^pikaos BovKUTieTs, Apiarapyos ^TparoXaos BoifKarieis, 10 AijKmv AdfiLOS^ Uevmv 'Ay^a-mv ^larvoi, "AXKtja-Tis. 'A d)v&, Keirai irapoi, 'Bivmva ^lo-tvSv. The inscription records the emancipation of a slave be- longing to one Lykos, an inhabitant of Arsinoe', by means of a fictitious sale to a deity*. In this case the deity is the ' Syrian Aphrodite of Phistyon.' The witnesses to the Deed are natives of Phistyon, or of Boukation. Phistyon is mentioned in the dedication published by Lolling: and the Upo^vXaKes of his inscription were probably concerned with the worship of this Syrian Aphrodite. With regard to the goddess, the epithet proves that we have not a native, but an imported cult, and a question arises as to the period at which the importation took place. Bazin " justly calls attention to the connexion with the East through ' Me'm. App. No. ii. The restorations are mine ; and in several places I have corrected Bazin's copy. See also Fick (Sammlung- CoUitz), 1428*. ^ Fick has : t\S)v AiVcoXSi/ . . . ' Fick corrects to Au[/c]io-[k]oi;. * Bazin: SFAIPOS. Fick: 2[jr]a[i]pos ; cf. W.-F. i5i=Sam»4. Co//. 1816: 'Sivaipos 'AKpuriov NauTrdicnor. ° Bazin : TArENOSOIKOrENHS. * Aa^ios might be taken as ' of Lamia,' indicating the native town of Lykon. But it is a common Aetolian name : cf. the SkAla Inscriptions. ' See p. 209 fol. * See p. 333, and authorities there quoted. ' Me'm. p. 3261 xiv] THE SYRIAN APHRODITE. 201 the alliance with King Antiochos III, at the beginning of the second century before our era. With the troops of the Great King, Asiatic ideas and Asiatic cults would readily penetrate the mountains of Aetolia. Nevertheless, we can- not acquiesce in the date suggested. Aet'olian free-lances had been roving about the eastern Wonderland for years before the alliance, with Antiochos. It was just because Aetolia was able to pour forth an inexhaustible stream of mercenaries at the moment at which there came a demand for them that the League rose to wealth and power^. From the death of Alexander, Aetolia and the East had been connected in this way; doubtless those who were so lucky as to return to their native highlands brought back many a strange faith along with the gold and jewels won in Egypt or Syria. We must, therefore, recognize the impossibility of thus fixing on a priori considerations the date of the introduction of this Syrian cult into Aetolia; all that we can say is that it probably came after the death of Alexander the Great ^. Of more immediate interest to the topographer is the addition of two names to the list of Aetolian towns. Neither Boukation nor Phistyon is known from other sources. Are they a real addition to our knowledge? It is, of course, practically certain that the find-spot of the inscription marks the site of the temple therein named. Where then was the town of Phistyon to which the temple belonged? If we examine the western face of the height overhanging the church we find, half-way up the slope, a few feet of wall, four courses high, built in 'irregular' style; fallen blocks at the foot of the hill are the only other vestiges. We search in vain for a continuation of the wall along the other sides of the peak; all trace of it has disappeared. The French map, besides making the mistake of putting the site . to the north-east of Kryoneru^, marks it as a palaiokastron, 1 Cf. Mahaffy, Greek Life and Thought, p. 7. ^ From the coins there seems to be clear proof that the Aetolian connexion with Syria began as early as the reign of Antiochos I, Soter. He issued three classes of bronze coins of European manufacture ; and, of them, two classes have in the exergue the jawbone of a boar, an Aetolian symbol. Cf. Holm, Gr. Gesch. iv. 264 ; BriT:. Mus. Cat. Seleucid Kings of Syria, p. 11, and Intro, xxiii. 3 And even Lolling puts it eight minutes north-east of the village. 202 CENTRAL AETOLIA [ch. i.e. as containing the ruins of an ancient town ; Bazin agrees with it^ We fail to discover any corroboration of that view in the single fragment of wall found on the hillside. It seems more likely that it belonged to the temple, forming part of a temenos enclosure, although this view also is not without its difficulties. Supported by an authority like Dr. Lolling^, however, we may confidently embrace the conclusion that we have here no evidence of the existence of a polls. We must then propose some other site for the name Phistyon, and the most likely one is the kastro of Sobonikos already described*. That town seems to have been unique among the Aetolian cities for the beauty and number of its shrines. We trace there the influence of circumstances that led to an architectural development not shared by the rest of Aetolia. For, in addition to the ruins already mentioned as existing in the neighbourhood of that kastro, there are others of a similar character, about fifty minutes to the north-east. We follow the before-mentioned stony path up the mountain to the village of Upper Prostovas*. Ten minutes below the village we see numerous well-wrought blocks, coming perhaps from an Hellenic building of which they are the only remains. To the south of Kato Pro- stovas there also exist the lower courses of a quadrangular Hellenic building of- good style. All these may have been connected with the town that stood near Sobonikos ; but we cannot positively affirm that this was the case, as the whole of this side of Mount Viena appears to have been thickly inhabited in ancient times. Still, the polls below Sobonikos must have been the most imporjiant place in the district, and, until additional evidence comes to light, the identification of it with Phistyon has more in its favour than any other. If ' Mem. p. 326 : ' d'insignifiants vestiges d'une ville etolienne.' ' Ath. Mitth. iv. 220. ' Tiiis is tlie proposal of Dr. Lolling. * MfiaaTo&as,' h.v(o and Karoa. For tlie history of the two villages, see Fiedler, Reise. i. 182. I am not sure that, in addition to the remains described, there is not a kastro of Haghios Elias above Prostovas. There is certainly a hill with a church of that name north of the village ; but I found no ruins near it, though the villagers regard it as an ancient site, unless indeed I or they confuse it with the Hdghios ElJas above Haghios Taxidrchis, lying to the north-east of Prostovds. XIV] SITES TO THE NORTH-EAST OF THE LAKE. 203 this be accepted, the identity of Boukation with the polis at Paravola seems to follow naturally ; for those are the nearest considerable remains. Analogous inscriptions, it is true, make it very plain that Boukation is not necessarily to be thought of as contiguous to Phistyon ; witnesses to Emanci- pation Deeds often belong to towns far distant from the scene of the ceremony. In default of any hint to the contrary, however, we may let the identification stand. We have, finally, a group of ruins at the north-eastern angle of the lake. Bazin appears to have been quite un- aware how rich in the remains of antiquity is this part of Aetolia, for he mentions^ only two of the four sites to be described. The first lies east of Prostov^s. A steep rocky eminence, bearing the usual title of Haghios Elias, rises above the hamlet Haghios Taxiarchis, the metropolis of Kephalovry- son. The hill is nothing but a mass of rocks disposed in oblique layers projecting above the surface in a series of knife-like ridges, over which we painfully scramble to the eastern end of the summit. There the rocks leave a small level space, which is occupied by the ruins of an Hellenic edifice. It is a quadrangular building,, lying east and west, thirty-three feet long and twenty-three feet wide, composed of fine blocks, but without any elaboration of style. In post-Hellenic times its northern wall was utilized to form the wall of the churcji of Haghios Elias, and another wall was run up the middle of the ancient building to form the south side of the church. The old west end then had a doorway broken through it; an apse thrown out at the opposite end completed the transformation and the ruin of the pagan temple. Nevertheless, we ought not to grumble ; it was not always that the Christian architects effected their recon- structions in this merciful manner. A remarkable feature of the site is the quantity of burnt bones found upon the slope below the temple ; the whole upper part of the bank appears to consist of these remnants of ancient sacrifices. The native explanation is that the hill was the scene of a battle with the very convenient Galatai. Coins are frequently picked up on the site' by the shepherds, and in the hamlet ^' Namely, Mokista (Mem. p. 324), and Berikos [ibid. p. 301). 204 CENTRAL AETOLIA. [ch. there survives a tradition of the discovery of a ' stone man,' with of course the usual yp6.ix)j,a7a (inscriptions) : but this, if indeed it ever existed, has long ago disappeared. On the north and east a great ravine separates the hill of Haghios Elias from the loftier spurs of Arabokephalon, on which the wretched village of Berikos^ stands. At the bottom of the ravine, the river Vasilikos flows from the bosom of Arabokephalon, falling into the Phidharis at the point at which that river turns suddenly southwards, below the village of Koniska. The kastro associated with Berikos is placed on a spur which projects to the south into the gorge, exactly opposite the hill of Haghios Elias and bearing north-east from it; a narrow saddle connects the spur with the mountain slopes leading up to Berikos. Kardharii^ is the modern name of the site. The remains are not worth the rough and tedious climb up the hill, but they give us a good idea of the rudeness of the Eurytanian settlements in the remote interior. The fragments of the wall enclosing the summit of the hill are composed of thin undressed slabs piled together without science. There was indeed little necessity for elaborate defences ; no enemy was likely to penetrate these wilds, and if he did his only way of approaching the fort was by the saddle, or up the eastern side of the hill,— both of them roads to certain destruction, if the garrison possessed a grain of resolution. The Vasiliko ravine on the south and' west rendered an attack from those sides an impossibility. So rough and impracticable is the region that it served the modern Greeks as a refuge from the Turks. Near a crag rising at the bottom of the gorge we see the ruins of a monastery known as the KaTa(pvyiov, the ' Place of Refuge.' At the eastern foot of the hill there is a spring, and near it an Hellenic wall of much better workmanship than that of the fortress. It was probably designed to guard the spring and the ascent to the fort from this side. From Berikos we descend again to the lake, to the village of Guritsa* which lies on the main road from Agrinion, four and a half hours distant from that town. Although quite ' MTre^iKos. It contains 303 inhabitants, and is tiie capital of tlie Deme ^ KapSapoij. ' Tovpna-a. xiv] GURITSA. 205 a small place, Guritsa is one of the most flourishing in this part of the country. Its prosperity is owing to its rich groves of lemon and orange trees, covering a level alluvial tract, a mile in length and nearly a mile in breadth, created at the north-eastern angle of the lake by the torrents of Guritsa and Mokista. Immediately west of this fruitful nook the hills come right down to the water in a rocky projection, similar to the spur of Sobonikos, from which it is separated by a narrow olive-covered strip of level ground. Below Guritsa the general direction of the lake is no longer east and west, but it makes a bend southwards along the foot of the steep slopes of Petrochori^- A quarter of an hour from Guritsa there is a small ^ monastery, called Myrtia frotn the myrtle trees which grow in profusion on the ridge. It is as picturesque and as dirty as any to be found in Aetolia. It possesses two churches in one, — that of the Panaghia, and one of much smaller dimensions attached to the southern side of the main build- ing and dedicated to the Archangels. The size of the Up6v, or chancel, in the larger church is noticeable; it is nearly half the length of the whole building. The interior of both churches is covered with fairly preserved frescoes, dark with age, but apparently of good workmanship, so far as an uneducated eye might judge. Some of them have been ruined by the Turks, whose musket-balls have also riddled the massive iron-plated door of the monastery. A small bell dated 1713 seems to be the only other antiquity^. Guritsa also possesses a kastro. It surrounds the ruined church of Haghia Paraskevi^, on a small rocky hill rising from the ravine on the south of the village. The church itself is to a great extent constructed of ancient blocks. The walls of the fortress are too much ruined to allow us to speak very definitely about them, but there is enough to prove the previous existence of a small polls on the height. On the south and west the abruptness of the descent towards ' See pp. 242, 256. ^ But the Christian monastery carries on the tradition of sanctity inherited from Hellenic times. To the pagan temples at Sobonikos, Kryoneni, and H. Taxi^rchis, there succeeded the Byzantine monastery in the Sobonikos kastro : and this in turn has been succeeded by the monastery of Guritsa. Cf. Ramsay, Hist. Geogr. of A. M. pp. 83, 414. ' 'Ayt'a TlapaaKevrj. 2o6 CENTRAL AETOLIA. [ch. the plain by the lake made elaborate works unnecessary: on the northern face of the hill we trace a few yards of wall built in fairly good 'irregular' style. The cemetery of the ancient town is found on the height by the roadside, just at the entrance to Guritsa from the west ; but hitherto nothing of value has been discovered in it, although the large slabs of which the graves are composed stimulate excavation. They are much sought after for making balconies. Certain scanty remains to be seen among the hills, about a mile to the north of the village, probably belonged to this polls. They lie on the southern side of a low ridge, but unfortunately the operations that have been undertaken with a view to the removal of the stones piece-meal by blasting them prevent our making out their original meaning; the plan has been completely obliterated. The size of the blocks, — three feet by one and a half, — and their careful dressing, seem to prove that we have before us the vestiges of a small but good building. One slab, standing on edge, measures seven and a half feet in length. The stones show sinkings for i 1 and i 1 cramps. About three miles south-east of Guritsa is the village of Mokista^, on the outskirts of which are remains of some interest. A large building, reminding us strongly of an English barn, stands by the roadside; it turns out to be a church, or rather two churches. We have already had an instance of this common practice of combining churches in one building; it gives rise to confusions, such as have gathered round this of Mokista. Bazin speaks of it as 'I'eglise des Saints- Archanges,' whereas Weil calls it the church of Sophia^. The truth is that the larger, northern, division is dedicated to Haghios Nikblaos, and the smaller to Haghios Taxiarchis, who -is of course Michael the Arch- angel. The ruins of an earlier Byzantine church lying close at hand belong to the Sophia of Weil. The whole site is called Palaiomonastero by the villagers. The interesting point about the churches is that they are almost entirely composed of blocks and slabs derived from an ancient temple ; large quantities of its materials are also " M^m. App. No. 8. Cf. Bull, de Corr. Hell. x. 188 ; Weil, in the Zeitschrift fur Numismatik, vii. (1880), p. 125. i H O in H s O H to 3' o s I4 o w o Bi a: u XI v] . TEMPLE AT MOKISTA. 207 scattered around half-buried, but nothing seems to remain in situ. By a lucky chance we recover the name of the deity to which the temple was sacred ; for, as we scan the walls in search of inscriptions, we find a large block walled into the southern side of the church of Haghios Nikolaos. It bears the following inscription, in careful letters i" high : — 'Apri/jLiTOS 'AyefiS^vos. The inscription occurs twice, at the' right and left hand near the upper edge of the stone : the final letters are much worn. On the lower part of the block there is an illegible inscription of a much later* date. Bazin^ takes the upper inscription to mean 'Artemitos, son of Agemon,' but the character of the stone, — a massive quadrangular block, — and the repetition of the words, prove that we have not to do with a mere sepulchral stele. Cousin ^ is certainly right in regarding it as a boundary stone, set up probably on the land belonging to the temple of Artemis 'Ayejuc^y ['Hye/io'r?;]. We learn from Antoninus Liberalis that Artemis 'Hye/no'v?; was also worshipped at Ambrakia, the key to Aetolian possessions in Western Hellas ^- A second important inscription is walled upside down in the apse of the larger church. It is too much worn to be made out completely ; in its present position it is impossible to improve upon Cousin's copy *, which reads : — ^TpaTayeovTOS tS>v A'itcoKoov $[i;]XraK]o[y Tov TlavToKemvos [ITXei/pcoi'] tou [IT] aratrcoXi- [/foty] e5o^e tols AItooXoTs, [«]"■[€ J . . . K^pare- os M . . . o[s xjprTyJo't/iors' etTTi npos to kJoivoi/ 5 Tav npo^evt . . TOV . . o . . . eafjio .... ' M^m. p. 324. * Bull, de Corr. Hell. I. c. Cf. Fick {Sammlung-Collitz), No. 1428''. ' Anton. Lib. iv. Cf. Polyainos, viii. 52: the daughter of Pyrrhos KaTevyev is to Upbv Tijf 'Hycjadvijs 'ApTe/uSos, at Ambrakia. Hesychios tells us that 'Hycjudvi; was also an epithet of Aphrodite. For Artemis Hegemone at Tegea, Paus. viii. 47. 6 ; Sparta, id. iii. 14. 6 ; Lykosoura, id. viii. 37. I. See Immerwahr, Die Kulte und Myihen Arkadiens, p. 157 ; Roscher, Lexicon. * Bull. p. 187. 2o8 CENTRAL AETOLIA. If the Pantaleon here mentioned is identical with the quondam Strategos of the League, Cousin is right in restoring UXevpmviov as the ethnic. We shall then also have approximately the date of the decree, for Pantaleon was in office for the first time in 186/5 e.g., again in 180/79 B.C., and for the third time in 173/2 b.c.^ In the fourth line Cousin suggests MtjAios, but there are many possibilities. At least five Byzantine inscriptions are found in the walls of the two churches, or lying on the ground near at hand. Bazin gives the text of some of them, but they are apparently of no great value ^. ' See p. 127, note 2. ^ Baz. M^m. p. 325 ; App. Nos. 9, 10. If the reading is correct, one of them teaches us that the church was rebuilt by a certain Kosmas, ■noBav Xa^elv KaBapaiv aimkaKifjiaTetv. CHAPTER XV. Central Aetolia. sites on the south of the lakes : the western section. When Philip of Macedon, in 219 b. c, prepared to invade the Paracheloitis he marched through the plain of Stratos in the direction of Metropolis and Konope'. After burning the former town he continued his advance on Konope, but 'the Aetolian horse rallied and ventured to meet him at' the ford of the Acheloos, which is about twenty stades before you reach the town ^.' Few towns in Aetolia are identified with a certainty equal to that with which Konope is placed at the modern village of Anghelokastron ^. The above passage from Poly- bios would suffice to establish the identification; but we may add others. That the regular ford of the Acheloos is in question in 219 B.C. is proved by reference to the account of the greater invasion of the following year. Then also, although his destination was different, the king crossed the river 'between the towns of Stratos and Konope*,' clearly at the ford already mentioned, as that used on the return journey is expressly stated to have been more to the north, near Stratos °. Strabo tells us that Konopa, as he writes the name ', was at first a mere village, and that it owed its foundation as a city to Arsinoe, daughter of Ptolemy I Soter, and sister and wife of Ptolemy II Philadelphos, king of ' Pol. iv. 64 : TToioufiej/os Trjp iropeiav i>s im MrjrpcmoXeas koI Kavairris. ^ Ibid, rrpbs Trjv tou irora/iou Sid^aaiv, fj Keirai irph Trjt irSkeas av 'urropiZp Koi TTOTafidv nva dvaypdei Kva6ov KoKoifievov, nepl 'Aptrivorjv ttoXu' AtrcoXiar. * AipiK6s. Cramer, ii. 69, calls it the Neschio, a name which he gets from Pouq., who also speaks of a second branch of the canal as the Primicos : Vqy. iii. 514, 516. ° So Leake, N. G. iii. 513. " Pouq., Foji. iii. 518, speaks of ' I'enceinte bastionnee d'Angdo- Castron . . . Les portes, les remparts et les tours de cette place, prfes de laquelle on voit un village de quinze families chretiennes, et le monastdre du Pantocrator, existent presque en entier.' I know not what to make O g 5 o o z o OS H tn < o K o z < O o xv] THE SITE OF KONOPE. 211 of Saint George, the walls of which are largely constructed of unbroken Hellenic blocks belonging to a fine building. The akropolis has only two other claims to notice. The first and most important is the character of the view from its summit. The range of the Zygos projects in such a way as completely to conceal the central and eastern portion of the ' great plain of the Aetolians under the peak of Arakynthos,' as Dionysios calls it ^ Only the extreme western end, even of the lake to which Anghelokastron gives its own name, peeps out from behind the green forest-clad spurs of the mountain. The prospect is towards ancient Agrinion in the north, and over the Acheloos into the plain of Stratos and the hills of Manina in the west. The second feature of interest is the number of artificial caves in the conglomerate sides of the hill. Each cave extends about four feet into the rock ; the roof is arched, rising three feet above the floor, which has a length of seven feet. The whole interior is coated with a greyish-white stucco, hard as marble, laid on very thickly in order to level up the natural irregularities of the coarse conglomerate. Traces of the old Aetolian city are principally discovered in the narrow plain at the northern foot of the akropolis ; the whole of the level ground on this side appears to have been included within the town. Among the remains is a cistern or shaft, with a diameter of four feet, supposed by the villagers to extend to an immeasurable depth, and to be connected with some passage communicating with the akropolis. It seems to be a rooted idea in the mind of the modern Greek peasant that every ancient citadel was furnished with its secret underground passage, leading often to a river or spring, sometimes to another citadel far removed^. A few yards to the west there is a subterranean building of the following description. It is a square chamber, measuring nine feet five inches each way. The roof is arched, the height of the keystone above the present level of the floor of this, unless he has confused this site with that of Palaiom^nina on the west of the Aspro. ^ Dion. Per. /. 431 : — . . . vno cTKomrip 'ApaKvvdov 'AvSpwv AItoK&v ttcBlov jiiya. * Oiniadai and New Pleuron are sometimes said to be thus connected ; Kal-damyle and Korone in Messeni^, and many others. P2 212 CENTRAL AETOLIA. [ch. being seven feet three inches. The structure of the arch exhibits a perfect symmetry. The keystone is eight inches- wide ; on each side of it are twelve regular courses, which gradually increase in depth from the keystone towards the ground, where the course is sixteen inches deep. In the seventh course from the keystone, not including it, we find remnants of iron nails at intervals along the four sides of the building, but none exist elsewhere. The whole interior is coated with a fine yellowish stucco to a uniform depth of one- eighth of an inch. Faint traces seem to indicate that it was once painted.. The lower courses of the wall on the south are inter- rupted so as to leave an. aperture 3' 3" high, and 2' 7" wide, closed from the outside by a large slab sunk so deeply in the earth that the peasants have not been able to remove it : it is generally agreed in the village that their failure is an indica- tion of something uncanny about the building, and of the existence of vast treasures under the stone. An entrance is now effected through a hole made by displacing one or two blocks of the roof The building is empty, save for a piece of Ionic half-column with nineteen flutings : it was evidently intended to be hermetically sealed, the hole covered by the slab being left for the escape of the workman who put on the stucco ^- Four large slabs, part of a stylobate, were discovered during the construction of the road which leads from the plain to the modern bridge over the Kyathos. They are now lying by the roadside, west of the hill, just before reaching the river on the way to the Railway Station. Three of them measure 3' 5" by 2' 3", and one shows a column mark two feet in diameter. The fourth slab is three feet square, with a column mark i' 5" in diameter. Em- bedded in the bank on the other side of the road is a piece of a Doric column, a httle over two feet in length, and i' 4^" in diameter, measuring from the hollow of the flutings. The larger drums which stood upon the first-mentioned slabs are probably still buried in the field above, for we find there certain pieces of wall which perhaps belonged to ' I could not learn that anything was discovered in the chamber, which was opened a short time before my visit. Probably it was first rifled centuries ago. ATTEMITOIATE MOT 12 a. APTEHITOL/^ OMjt^j'^ 12 b. 13- 14. 12 fl, 6. INSCRIPTION FROM m6kista. See p. 207. 13, 14. INSCRIPTIONS ON CLAY SLABS, ANGHEl6kaSTR0N. Tofacep. 213. XV] INSCRIPTIONS AT ANGHELOKASTRON. 313 ^ the small shrine, or whatever it was, from which came the ' fragments lying by the roadside. Comparatively few small objects are found, owing to the fact that for some reason the plain which contains the anti- quities above described is not under cultivation. I saw in the village two slabs ^ of hard red clay, three inches thick, each containing a sipgle word stamped upon it near one end. The first reads : — Avfravta. On the second slab the word, by a common oversight, is reversed : — ApfflVOiWV, Bazin also reports an inscription : — MOKPATEIA PYKAEIAA AAMinnAS which he translates:, ' Mokrateia, Rykleida, Damippasl' Obviously it must be restored thus : — AoL^iioKpaTiia Ev\pvK\ei8a. Aa/iiiTTras. As usual, Bazin gives no details about the stone. It is possible that Konope, the name first borne by the town, points to the unenviable reputation that still clings to the site by reason of the low marshy ground between the lake and the Acheloos ^. There is something rather myste- rious about the change of name, from Konope to Arsinoe, nor has the truth about it hitherto been reached *. It is not ' I was assured that the slabs came from different tombs, but the circumstances of their discovery seemed inconsistent with this. They are in the possession of Xpioroj Kp«e\i;s of Anghelokastron. ^ Mem. p. 319 ; App. No. 6. Cf. Fick {Sammlung-CoUitz), xap&^, where the first word is restored as 'Ep/ioicpdTein. ° Kiij'tB\p' = mod. KovvovTri, mosquito. * I cannot guess why Lolling should have imagined Konope and Arsinoe to have been two separate places. He puts Konope at 214 CENTRAL AETOLIA. [ch. sufficient to repeat the statement given in the pages of Strabo ; yet usually no attempt is made to specify the nature of the connexion existing between Aetolia and the Egyptian king, or to exhibit clearly the meaning of such foundations as this of Arsinoe. Two considerations lying very close at hand make against the reality of the supposed connexion with Egypt. Firstly, a neighbouring town also bears the name of a foreign potentate ; secondly, its name, — Lysimacheia,— is that of a husband of this very Arsinoe who figures in Strabo's account. For Arsinoe, before she became the sister-wife of the second Ptolemy^, had been taken in marriage by Lysimachos king of Thrace, in 299 b. c. ^ We suggest, therefore, that Konope was strengthened by the Thracian king, not by the Egyptian queen. Lysimachos performed precisely the same service for a city more famous than this Aetolian village, — for Ephesos, which he restored and called Arsinoe in honour of his consort ^. It is easy to understand how from the new name there should have grown the mistake of supposing that the queen herself was to thank for the improvement of Konope *. Lysimachos fell in the battle of Koroupedion, in 281 b. c.^, so that his connexion with Aetolia must belong to the interval between 299 and 281 b. c. We can fix the date more nearly when we recall the fierce warfare carried on between Lysi- machos and Demetrios Pohorketes ^- Seeing that Demetrios Anghelokastron, and Arsinoe ' nordlich unter Ihm ' in the direction of the lake (Mtlller's Handbuch. iii. 139). Nothing is to be made of the fact that Polybios uses the two names indiscriminately (Konope, iv. 64 ; V. 6, 7, 13. 'Apo-ii'dij, xvii. 10 ; 'Apo-ivoia, xxx. 11). Steph. Byz. has 'Apo-tTOij, TToXis AiTcaXiar, whereas he makes Kcoixbtt)) a TtSKis 'AKapvavlas. Bursian, Geogr. Gr. i. 135, note 3, is not quite convinced of the identity of Konope and Arsinoe ; but surely we cannot go behind the express statement of Strabo, whatever we decide as to the value of his words with regard to the historical question of the connexion with Egypt. ^ Paus. i. 7. 1 ; Droysen, Gesch. der Epig. i. 265 ; Holm, Gr. Gesch. iv, 237. ^ Paus. i. 10. 3. ' Str. p. 640. Staph. Byz. in voc. 'E-Z)!W.ii. 311. ' Gesch. der Epig. ii. 327 : ' man wflrde, da sich in der Nahe auch ein Lysimacheia befand, vermuthen kSnnen, dass Arsinoe als Gemahlin des Lysimachos die nach ihr genannte Stadt stiflete, aber Strabos Ausdruck verbietet es ; als Ptolemaios Gemahlin, also nach 267, hat sie die Stadt gegrlindet.' * Str. p. 460 : KTiana 6' VTTrjp^fv 'Apaijirfijf rijr IlToXefiaiov toB 8evT£pov yvvaiKos Sjua Koi d8e\AAAKP0Y ^oKaKpov. Graves are discovered in the fields along the road to the east, in the neighbourhood of Mataranga. In the pavement of the church of Saint George in that village there is preserved the upper part of a stele ornamented with two rosettes, above which Bazin deciphered the letters ^ — XIAA XPiriNO which he interprets as ' Chidas son of Chrion ' ; but we should restore — 'A p\ xt5a[jtioy Bazin gives no information as to the state of the stone in his time, and the inscription has been all but obliterated by the feet of generations of worshippers. Other reported inscriptions have shared the fate of most such discoveries, having been broken up for building purposes. Now that the site of Lysimacheia is known we can interpret the expression of Strabo, who speaks of it as ' near Olenos ^.' It is clear that we cannot take him literally, for Olenos is below the kastro of Irene, separated from Lysimacheia by the breadth of Arakynthos, a distance of at least ten miles. We need not, however, put aside Strabo's words as a mere stupid blunder. The key to his meaning is to be found in the fact that Pleuron, as well as Olenos, is brought into relation with Lysimacheia ^. For Lysimacheia, in addition to being what Strabo calls ' near Olenos,' is stated to be ' between Pleuron and Arsinoe.' We have already sufficiently described the ^ Mem. p. 320; App. No. 7. Cf. Fick {Sammlung-ColUts), 1428 b, where the first word is restored as MaxiSaj. The letters XIAA are still with difficulty to be read on the stone. ^ Str. p. 460 : Rnra he ttjv fdroKiav ^v 'QXcvos , . . ^v 8e Koi Ava-ifidx^M 7rX)j(n'ov. ^ See p. 134, note i. 224 CENTRAL AETOLIA. [cH. two most direct routes leading from the maritime plain into Central Aetolia'. In both cases Lysimacheia marks the point at which the road enters the central plain, — the road through the Kleisura enters it on the west, the path over the Zygos on the east, of the town ^. Now, the main path crossing the ridge by way of the ancient Elaos starts from Pleuron. Pleuron and Lysimacheia, therefore, guarded the two ends of the mountain road connecting the two Aetolian plains, and there was thus a very real and close connexion between the two towns ; in which connexion Olenos also shared, from its proximity to Pleuron. It is that relation which, half clumsily, half inaccurately, is expressed by Strabo. Thus we have further light thrown upon the work of Lysimachos ^. In strengthening Arsinoe and Lysimacheia he fortified two important lines of communication between Central and Southern Aetolia, — precisely the work that was subsequently completed by Attalos in the fortification of Elaos. The instinct which pitched upon the site of Konope was more sure than that which selected those of Lysimacheia and Elaos ; and the future may still further justify it *- A serious difficulty confronts us when we endeavour to assign their ancient names to the two Aetolian lakes. We have, in the first place, the plain statement on the part of Strabo, that Lysimacheia was near a lake called in his time after that town, but formerly named Hydra*. If, as we should naturally argue, the town gave its name to the nearest water, we must identify Lake Lysimacheia with that now known as the lake of the Apbkuro or of Vrachori, for Pappadhatais overlooks its western end. The kastro of ' See p. 146. We are, of course, disregarding the most easterly route of all, that leading from Naupaktos behind Mount Chalkis over the Phidharis, and debouching into the central plain near Kaludhi (cf. p. 240). The fourth, or most westerly route, is that through the Sreca of the Acheloos, by way of Stamna and Angheldkastron (see p. 146, note 2). ^ Cf. Leake, N. G. i. 122. ' It must be noticed how admirably the attribution of the two founda- tions to Lysimachos harmonizes with what we know of him. For Lysimachos was the most brilliant strategist of his age. * See p. 51. ° Str. p. 460 : irpbs TJj Xi'/ivn ''B "vv jiiv ti.v(Tifiax^ia irpirepoii 8' 'Ydpa. XV] THE LAKES IN CENTRAL AETOLIA. 225 Pappadhatais has no natural connexion whatever with the lake of Anghelokastron, although the contrary is often asserted by those who, to avoid a difficulty, wish to give the name Lysimacheia to that lake ^. Still, the identification with the lake of the Apokilro is apparently not borne out by Polybios, who, in describing the march of Philip, mentions only one lake, calling it Trichonis ^. Whatever else may be asserted, this stands fast : by Trichonis Polybios could only mean the larger and most easterly of the two lakes, that of the Apokuro. It is quite possible that at different times the same lake bore different names. A parallel for this could be drawn from the modern history of the lakes themselves. Pouque- ville, for example, preserves the name Siidi, which seems now to be forgotten ^. We might, therefore, without much hesitation, imagine that the lake of Vrachori was first called Hydra, afterwards Lysimacheia, and finally Trichonis. This, however, does, not extricate us. Strabo asserts positively that the lake bore the name Lysimacheia in his own day, though the town itself was a ruin. We should, therefore, be compelled to imagine, either that the name Trichonis came before that of Lysimacheia, and was in use for a com- paratively brief period, or that different parts of the lake bore different names. Against the former supposition we must set, firstly, the fact that Trichonion still existed as a city in the time of Strabo ; and, secondly, the difficulty of believing that a name derived from a place of such importance should ever have been discarded in favour of one derived from Lysimacheia, a decaying or even already deserted town. If we accept the second hypothesis, that different names ^ As Leake, N. G. i. 153 ; and Lolling, who says : ' Die Hyria wurde nach der Stadt Konope, spater nach der Stadt Lysimacheia benannt.' I. Mailer's Handbuch. iii. 139. ^ Pol. v. 7 ; xi. 7. ' Vqy. iii. 513. He writes: 'nous eumes la vue du lac Trichon, que las modernes appellant Ozeros, lac da Vrach6ri et Soiidi, denominations qui repondent a celles de Lysimaque, d' Hydra ou Hyria et de Trichon, qui ne servirent jamais qu'a designer les diverses parties d'un saul et meme lac' His map gives the name Sadi to the eastern lake, so that he evidently looks upon that of Anghelokastron as its western end, dis- regarding the marsh of Ali-bey. Although the name Ozer6s (a generic term) might be applied to the lake, it is restricted iii usage to the Akarnanian tarn in the plain of Stratos, at the foot of Lykovitsi. Q 226 CENTRAL AETOLIA. [cH. were attached to different parts of the lake, we must still inquire what was the name of the smaller water, that of Anghelokastron, or, as Leake calls it, the lake of Zygos \ There is equal confusion here. From Antoninus Liberalis, who takes his narrative from Areus the Lakonian and the Metamorphoses of Nikander, we learn that the lake of Anghelokastron was first called Konope,— as indeed we should expect from its proximity to the village Konope, which afterwards became Arsinoe. The Aetolian youth Kyknos, who threw himself into the lake ^, was turned into a swan by Apollo ; but his mother, whom Antoninus calls Thyrie, in despair drowned herself in the lake, which was thenceforth called Kykneia ^. Ovid, however, calls both the mother and the lake Hyrie *- For the lake of Anghelokastron, then, we have the following series of names : — Konope, Kykneia, and Hyrie ; not one of them seems to be known to Polybios or Strabo, who are both to all appearance ignorant of the very existence of the smaller lake. For the lake of Vrachori we have likewise three names : — Hydra, Lysimacheia, and Trichonis. In addi- tion to these variants, the possible connexion of Ovid's Hyrie with Strabo's Hydra, as well as with the lake Ouria ^ near the coast, constitutes a further difficulty. The name Hyrie is almost certainly due to the confusion which I have just suggested. Thyrie, which occurs in Antoninus only as the name of the mother of Kyknos, has been changed by Ovid (or by the author from whom he derived his narrative) into Hyrie, and has been extended to the lake in which she was drowned, — the change being 1 N. G. i. 124. ° Anton. Liberalis, xii : As rqv Kavamfv \eyoiiivr)v XI/jlviiv. ' Ibid. QvpiT] T) firjrqp «carej3aX«' eauT^v els t^i/ avn)i>' ixelva Xiimriv . , . Kai r) Xifivr) iieravondtrdri koi iyhero Kvki/cii;. He adds two small points : — Kal TToXXoi iv TJ] &pa Tov apoTQV ivravda aivovTcu kvkvoi.' »r\i;(rioi» fie xeirai Kal ri Tov ^vXiov trrjfia, * Ovid, Met vii. 371 : — Inde lacus Hyries videt, et Cycnela Tempe which latter must indicate the Kleisflra. Ibid. 380 : — At genetrix Hyrie, servati nescia, flendo delicuit ; stagnumque suo de nomine fecit. Adiacet his Pleuron. » Str. p. 459. xv] THEIR ancient' NAMES. 227 helped, if not originated, by the existence of the somewhat similar words Hydra and Ouria as names of Aetolian lakes ^. Further, the name Kykneia, which is apphed to the lake by Antoninus, does not bear the stamp of a genuine appellation ; that also is a poet's fiction, introduced to round off the story of Kyknos in the appropriate and traditional manner. We unhesitatingly, reject its claim to appear upon the map of Central Aetolia. We are thus left with Konope as the only genuine title of the lake of Anghelokastron during the historical period, — if it had a name at all, or indeed any separate existence as a lake. For the failure of Strabo and Polybios to notice the fact of there being two lakes, and still more the physical structure of the basin, are enough to make us ask whether it is not probable that in ancient times there was no division between the two sheets of water ^. Hydra, the original name of the larger lake, may have belonged primarily to some town on its shores. Possibly it was the old name of Lysimacheia ^ ; if so we can readily understand how the lake should have borne both those names at different epochs. The two names Lysimacheia and Trichonis may have been in existence contemporane- ously, the former being applied to the western, the latter to the eastern end of the lake. However that may be, as Lysimacheia fell into decay, the name Trichonis won its way by little and little, until it was the only one known to Polybios. Strabo, however, in his authorities found the name Lysimacheia recorded, and copied the information. This seems to be the only hypothesis by which we may reconcile the conflicting statements of poet, geographer, and historian. ' It is quite unjustifiable to change the "YSpg of Strabo (p. 460) to 'Ypi'g, as Bursian (Geogr. i. 135) suggests. It is not for Ovid to rectify Strabo, but it is Strabo who convicts Ovid of geographical confusions. ^ See pp. 49, 272. ' That is, when the place was still a Ka^rj, before its re-foundation by the Thracian king. Or Hydra may have been an old name of Trichonion. See p. 133, note 5. We may express the sequence of names thus : — (Konope) —Hydra ; (Konope)— Lysimacheia-Trichonis ; (Kon ope) — ^Trichonis. Q2 CHAPTER XVI; Central Aetolia. HE I. It is with the eastern section of Central Aetoha that pur topographical difficulties really begin. In Lower Aetolia the problem is merely one of arrangement ; we have a certain number of sites, and a certain number of ancient names to be distributed among therii : and no site is left anonymous. Here the case stands otherwise: sites abound, but names are wanting. Scanty, however, as is our knowledge of this most important part of the country, we are fortunate in still having Polybios ^ for our guide ; the valuable services ren- ' In Turkish times the kasd of Karlili, which embraced all Akarnania and most of Aetolia, was divided into four parts : — Valtos and Xeromeros on the west of the Acheloos ( = Akarnania) ; Vlochos and Zygos on the east of the river. The line of division between the two latter districts was the northern shore of the lake of Anghelokastron, which was thus entirely included within the district of Zygos. On the lower Euenos Zygos bordered on the Venetiko, the district of Naupaktos. The eastern section of the central basin was called Apokuro, which wholly included the lake of Vrachori, being severed from Kravari by the middle Euenos. Kr^vari constituted a separate district, but Apokuro belonged to the kaza of Karpenisi. North of the district of Vloch6s came that of Agrapha. See Leake, N. G. i. 124. Leake suggests that the name Ap6kuro may be a corruption of the ancient name Kourion, as part, at least, of Mount Arakynthos (Mount Zygos) was called (cf. Strabo, p. 465 : roC Spous roO Koupiou toS iirepKci- fiivov rrjs llXevpSivos. See also id. p. 451). Ap6kuro would thus = the region 'behind Mount Kourion,' i.e. N. of the Zygos. " Cf. Palmer, Gr. Descr. p. 461 : ' O si Polybium integrum haberemus, quanta lux in Graeciae descriptione nobis affulsisset ! ' THE EXPEDITION TO THERMON. 229 dered by him in the Paracheloitis will be found to be repeated, and even surpassed, in Central Aetolia. Our knowledge of the towns round the eastern lake is, in fact, almost entirely derived from his account ^ of the Mace- donian invasion of 218 b. c. Half the Aetolian army being then absent in Thessaly with Dorimachos, the Akarnanians seized the opportunity to prevail upon Philip to strike an unexpected blow at the capital of Aetoha. Philip set out from Limnaia with his troops in light marching order ^ and with guides selected by himself. Starting in the evening the force made sixty stades. After a short halt ^ to enable the men to snatch a hasty meal, the king put them in motion in earnest, ' and marching right through the night* arrived, just as day was breaking, at the river Acheloos, between the towns of Stratos and Konope, being anxious that his entrance into the district of Thermon should be sudden and unex- pected. ... He continued his march without interruption ; and after crossing the Acheloos advanced rapidly upon Thermon, plundering and devastating the country as he went^, and marching so as to keep Stratos, Agrinion, and (the) Thestieis on his left, Konope, Lysimacheia, Trichonion, and Phytaion on his right. ' Having arrived at the town of Metapa, which is on the borders of the Trichonian lake, and close to the narrow passage along it, about sixty stades from Thermon, he found it abandoned by the Aetolians, and occupied it with a detach- ment of five hundred men, with a view to securing his entrance into and exit from the defiles : for the whole shore of the lake is mountainous and rugged, closely fringed with forest, and so afibrding but a narrow and difficult passage. He now arranged his order of march, putting the mercenaries in the van, next them the lUyrians, and then the peltasts and men of the phalanx, and thus advanced through the pass ; his rear was protected by the Kretans, while the Thracians 1 Pol. V. 6-14. ^ TO iroKv rrj! imoirKfvfjS dnoBffUvovs d^avovs as Trapa(TKevd^eiv npos ava^vyrpi. ' 0pa)(i 8tavmrav(Tas. * (ruye;^£s vviiT(nropriy Taxurra biavvdm tcls fiuirxoipios bia to irpo(r8oKav tow AiT(oKovs e^dyjfetrdai Trjs avpaylasy irKTrevovras rats oxvpoTrjtri Toiv roircov. Ibid, ajia be tm KiVTJtrai rrfV oipayiav iiri^oKov evBias eis rov Qeppav Kai irpo *"' KaraarpaTtmeievaas jrcpl '&.aiv&'in)v eircptive rf/v ex"/*^"'?'' W^P""*' ^ Ibid. fTToteiTO rfiv itopeiav irapa rltv 'Ax^^V"^ '"" '''^' ''^'' ^'poTOK. dia|3ar Bi TOP TTora/idy k.t.\. ' Pol, xi. 7. * Steph. Byz. ^ircuov, 7r6\is AlraKias. HoXu/Stos ivSeKart/. To idviKov *UTaIoi. For Ellopion, see p. 261, noie i. 232 CENTRAL AETOLIA. [cH. depend solely upon the indications given in connexion with this invasion, and it remains unaffected by the particular theory adopted with regard to the position of the capital and the road by which the Macedonians reached it. As to Trichonion, the next city in order, scarcely any doubt is possible. From the fact that Strabo took it, along with Stratos, to define the extent of the central Aetolian plain \ and from the fact that it gave its own name to the principal Aetolian lake, we infer that Trichonion was the most im- portant town in this region. The site of Gavalu ^ corre- sponds to this inference, as it is situated in the richest part of the country on the south of the lakes, a little to the west of the point at which the level ground ends and the spurs of the Zygos come right down to the water. The importance of the site is not strategic, but arises entirely from its central position with regard to the finest land in Aetolia. The village lies about five miles east of Pappadhatais, on the western slope of a low ridge running from east to west, and connected with the hills leading up to the Zygos on the south. From the northern .foot of the height, the plain, richly covered with maize, tobacco, vines, and olive trees, extends to the margin of the lake of Vrachori, a distance of about two miles ^. Scarcely anything survives of the fortifications ; here and there only, along the sides of the hill, a few stones enable us roughly to determine the line of the enceinte. It is only at the eastern end of the ridge, where it is less steep than on the west, that any considerable fragments of the wall' are visible ; but even there we find only disjecta membra, from which it is impossible to carry away any complete idea of the system of defence. There is, however, enough to show that a gateway existed at that point, and there is some trace of a projection at the angle. The style of masonry is somewhat " Str. p. 450. See p. 57. ^ Ta^aKov. Capital of the Deme MaKpweia. It has about 700 inhabi- tants. ' So Leake (N. G. i. 155) writes : ' the occurrence of its name (Trichonion) after that of Lysimachia, among the cities on the right of Philip in his progress towards Thermus, places it beyond a doubt towards the south- eastern extremity of the plains, where GSvala, in a fertile district on the -southern side of the LakeofAp6kourD,seems.perfectly to correspond to the data.' xvi] THE K ASTRO OF GAVALU. 233 rude, reminding us of that of Ithoria, and of the earlier work in Old Aetolia. The general impression gained is that these ruins are more ancient than any to be found along the southern shore of the lakes ^ ; if they are those of Tri- chonion this is in agreement with the early importance of that city. Although, the site is thus of little value as a specimen of Aetolian fortification, it has yielded more than any other in the way of small remains ; probably much is hidden under the soil at the foot of the kastro. The white walls of the church of the Panaghia on the top of the hill are conspicuous even from the road on the northern side of the lake. It may stand on the site of an ancient shrine, as fragments of Ionic columns are found near it. Similar fragments are also unearthed near the akropolis wall on the east, below the church. They are fifteen inches in diameter, of coarse stone covered with a very white thick stucco. A lion's-head gar- goyle to be seen in the village probably came from the same building. Many tombs are opened on the plain at the northern foot of the hill ; they yield small ornaments, and vases with rude incised designs. Five inscriptions have also been discovered on the site ; but as they are all, with one exception, from grave stelai they are of little importance. The exception is a dedication, of fairly good period. One of the grave-stones is Roman. They thus indirectly testify to the existence of the town throughout the period of the League ; and this again is in harmony with its identification as Trichonion. The inscriptions are the following :-^ (i) On a plain slab of sandstone %' x i', cut at the bottom for insertion in a basis : the upper edge of the stone is bored in order to receive the anathema. Letters good, i^" high ^- Tpmiiig (2) Veined white limestone stele, decorated with oak leaves and rampant animals, now lying in two pieces at the spring on the northern side of the akropohs. Letters carefully cut, ' See p. 133, Mote 5. ^ This inscription, as well as No. 2, first published by Cousin, Bull, de Corr. Hell. 1886, p. 189. 234 CENTRAL AETOLIA. [ch. in a sunken band. There is nothing wanting, though the first letter is shghtly mutilated. ALLIA • C . L ■ LENA (3) On plain stele of grey stone, now walled in above the fireplace of a house near the aforesaid spring. Above the inscription there is an incised equilateral triangle, with traces of red paint; but the colouring may be a modern addition. Letters i^" high, rude and coarse. AAESIA? (4) Among the vineyards, a quarter of a mile east of the akropolis, on the left hand as one goes to Burlesa. Same material as No. 2 ; namely, a hard white veined stone, which seems to have been imported. Letters fair, i|" high ; second word very faint. The stele is about 18" high. ' flipeXloivos 'Av6iiTira9. (5) A worthless fragment which cost me much trouble to re-discover. It is a stele, once preserved at the ruined church of St. John, a kilometre from the kastro, in the direction of the lake. Here it was seen by Cousin ^ Since his visit it has been removed to the church of Haghia Paraskevi at' Kalphenikion, and ruined by the addition of rude figures. I could only decipher : — XA'- -Er^Y Cousin's copy is probably more trustworthy. He read : — XAPI-ENOY Xapdf\ivov. Trichonion gave birth to a considerable number of pro- minent men. It is chiefly identified with the turbulent family of Nikostratos, who gained notoriety by some underhand ' I.e.' Dans les ruines d'une eglise au milieu des champs.' T P-n.| At AN EOHKE 15 «■ AN01FPA t 15 1>- Xll/a-CL'le N a 15 c. 15 a, b, C. INSCRIPTIONS FROM GAVAl6. Tofacfp. 234. xvi] HISTORICAL POSITION OF TRICHONION. 235 attempt upon the Boiotian Federation^- Dorimachos^ his son, twice Strategos of the Aetolians, covered Aratos with shame at Kaphyai ^ and himself with dishonour at Dodona *. The physically infirm Ariston ^, and the headstrong Skopas, both of them kinsmen of Dorimachos, kept alive the violent tradi- tions of the house. Dikaiarchos and Thoas, two brothers who in turn held the supreme office of the nation, had to their credit the chief share in the negotiations with Antiochos, which led to the ultimate ruin of both Aetoha and the Great King ^. Nikandros and Proxenos, the former a fellow-exile of Polybios in Italy, complete the list of Strategoi from Trichohion. No other town could boast of so long a hst or of so commanding an influence upon the politics of the nation. Alexander, who threw himself upon Philip's rear-guard after the sack of the capital, was also a Trichonian ; many other natives of the town are known to us from inscriptions. We get the idea that, owing to its favourable situation, Trichonion early became the home of wealthy chieftains whose riches enabled them to win great names for themselves, and we must regret the more that so little has survived of the city whose sons played so large a part in the history of their country. The next place mentioned by Polybios is Phytaion. It is with this name that divergence among the topographers begins, for some put the town on the north, others on the south of the lake. Yet there is in Polybios no hint of any change in the direction followed by the army. Konope, Lysimacheia, Trichonion, Phytaion, follow one another without a break ^, so that we are justified in searching for ^ Pol. iv. 3; ix. 34 : T(' be AdTTo/Soj icai NiKoorpaTor j oh TrjV t&v Uaii^ouonav navrfyvpiv, elprjvrjs ov(rt]S, irape(m6vhria-av, 2Kv6i>v epya Kai TdKarav eiriTeXovvreg &v oiScv treirpoKTai toIs ScaSe^afieUois ; " Pol., iv. 3, characterizes him as TcXrjprjs AtrcoXiK^s 6pp,ris koI liKf ovinias. He was Strategos in 219/8 and 210/9 ^.c. ' Id. iv. II fol. * Id. iv. 67. ° Id. iv. 5 : 8ia nvas triopanKas a(r6ei/eias aSivaros i)v jrpag TToXe/UK^i/ xp^iav. ^ Id. xxi. 31 : eWt h' cmo piv rris 'Atrtas irvfia-aVTiS &6as Kai AiKaiapxos, (IttA fie rrjs 'Evpomis MevetrTas Koi AaftoKpiTos frvverapa^au Tois S^fXous K.r.\. '' Id. v. 7 : Trap^ei 8' ck pev tvmvvpav airokmav STparov 'Aypiviov eeoTKir, eK 8e be^iSiv Kdii/onriji' A.v(Tipaxaav Tpix^viov ivraiov. 236 CENTRAL AETOLIA. [cH. the site of the last member of the series in the direction which we have hitherto followed. After leaving Gavalii we pass through Burlesa ^ and, still going eastwards, find the foot-hills of the Zygos advancing farther into the plain; at the same time the lake makes a bend towards the south-east, so that the plain at last entirely vanishes, and the road is cut in the rocks or supported on buttresses along the edge of the water. Five or six miles from Gavalu we emerge upon a level recess at the head of the lake. The last member of the hills on the west of this plain projects towards Lake Trichonis in a long high steep ridge, on which we see the village of Palaiochori ^. The main road, which we have been following, does not ascend to the village, but keeps along the eastern foot of the ridge, in the direction of Makrynii. The summit of the hill of Palaiochori forms a very narrow platform, which on north, east, and south-east, falls abruptly to the plain. On the west it descends more gently, yet somewhat steeply, to the stream that separates it from other similar projections springing from the main hills on the south. Besides being built actually on the line of the western wall, the village is probably composed entirely of materials taken from it. The steepness of the hill on the remaining sides, especially on the south-east, made it un- necessary to have more than a straight line of wall along the western brow of the plateau, with a short return at either end running across the ridge. The wall seems to have had square towers ; at any rate the foundations of one appear under a modern house at the southern extremity of the line : it projects fifteen feet from the curtain. The defences nowhere rise higher than two courses. They are remarkable for their extreme solidity of construction, and for the great size of the blocks employed. The breadth of the wall is about nine feet : the style ' irregular Hellenic' No remains are found within this narrow akropolis, which is now planted with corn. The ancient town probably lay on the slope below it, in the direction of the western ravine*. Antiquities ' MTToupXeVa. ^ HaKaioxiipiov. ' The ridge does not fall immediately into the lake, being separated from it by a somewhat lower projection or spur. This also seems to have been occupied in ancient times. xvi] PHYTAION AND THE MACEDONIANS. 237 have in fact been discovered near the spring, or Kephalo- vrysis, to the north-west of the village. Among them were a torso, and a marble hand grasping some round object, supposed by the peasants to be an egg, but probably in- tended to represent a pomegranate or other fruit. I found it impossible to trace these objects farther than the Eparchion of Mesolonghi. These ruins must be identified as those of Phytaion ^. In his account of the first invasion Polybios merely mentions the name of the town, not a word being said of either its occupation or its destruction : in this respect it resembles Trichonion, Lysimacheia, and Konope. Doubtless all those towns were, like Metapa, found deserted ; and Philip, as he approached, would send forward a detachment to occupy them until the column had passed. For obvious reasons there was no need to leave a force in possession ; nor was there time to destroy their works. The case of Pamphia and Metapa was different; owing to their situation at the two ends of the defile, it was absolutely necessary to occupy them with a garrison ; and subsequently, if possible, to raze their fortifications, in order to set the seal to the destruction of the capital by dismantling the fortresses that covered it. On the occasion of the second invasion, however, there seems to have been a change ; in some way, Phytaion, along with the hitherto unheard-of town Ellopion, figured prominently in the operations of the king. The fact that Phytaion comes immediately before Metapi suggests the nature of those operations. Philip had un- covered the approach to Thermon as effectively as was possible in the short time at his disposal in 218 b.c. In 206 B. c. he completed his work by demolishing the outer zone of defence. The appearance of the walls themselves confirms this theory. The immense solidity of the enceinte of Palaiochori is unmistakable ; we find no such masonry elsewhere on this side of the Trichonian lake. It is, in fact, only in parts of the wall at Vlochos that we see anything to resemble it. Walls like these were not to be razed without appliances, and time to use them. Phytaion, therefore, escaped destruction in the flying inroad of 218 b. c, but the ' So Bazin, M^m: p. 321. Lolling, in accordance with his previous identification of Pappadh^tais as Trichonion, puts Phytaion at Gavalu. 238 CENTRAL AETOLIA. [ch. more deliberate invasion made twelve years afterv^^ards saw it permanently crippled, at any rate as a fortress. An inscrip- tion from Delphi gives us the name of one Panaitolos, a native of Phytaion, and Strategos of the League some time during the first half of the second century ^ At this point difficulties crowd upon us. We have still to locate four towns,— Metapa, Pamphia, Thermon, and Akrai. That is the order in which they appear in the narrative ; but it is not their real order. The true sequence, from west to east, is Akrai, Metapa, Pamphia, and Thermon. Nevertheless, their identification, in the absence of evidence from the sites themselves, must depend solely upon the order in which the names occur in the story of events. If, then, we can fix the position of Metapa, the first link in the chain of narration, it should not be difficult to locate the other towns, seeing that in two cases Polybios has furnished us with precise numerical data in addition to the- other niceties of his description. We are here, in fact, face to face with the central problem of Aetolian topography, for the secret of the true position of the capital might almost be said to have died with the nation. The natural difficulties which, in men's imagina- tions more than in actual fact, fenced round the 'akropolis of Aetolia ' have reappeared in the far more formidable barriers of merely partial investigation or prejudiced judge- ment. How baneful in their effects these can be is shown by the examples of Leake and Bazin, who both failed to solve the problem, — the former owing to the incompleteness of his explorations, the latter owing to his preconceptions. It is an interesting commentary on their efforts to find how near the truth Bursian comes ; so near that Dr. Lolling, the next inquirer in the field, reads the riddle. In many ways, ^ Baunack {Samm. Coll.), 185^ = W.-F. 189. Panaitolos dates also the inscription Baunack. 1729 = W.-F, 64. He was Strategos at some time between 170 and 157 b.c. Bechtel (Samm. Coll. 1511) gives a fragment from Lrjuapxla, near Neon-Thronion, in the country of the Opountian Lokrians ; it runs ['App^ovros . . .] Qpoviels ebuxav ... cos AiroiXrai IK *u[t(u'ou avT&i K]ai iK.y6vois rhv [jrayra ^povov] npo^eviav k.t.\. This belongs to the same period. Cf. Girard, De Locris Opuntiis, p. 48. XVI] THE PROBLEM OF CENTRAL AETOLIA. 239 however, Lolling fails to be a true successor to Bursian, and, although he correctly identifies Thermon, his other sugges- tions are inferior to those put forward by the writer of the Geographic von Griechenland, — nay, the very identification of the capital is in defiance of the logical results of LoUing's own scheme of topography, as revealed in the sites suggested for the remaining towns of this district. Nor, again, did Lolling make the slightest attempt to exhibit the process by which the identification was reached, or to establish its truth in the face of the difficulties surrounding it. That there are difficulties it would be idle to deny. They are not here due to the ravages of centuries and the destruc- tion of all points of contact between the literary record and the face of nature, but to the very multiplicity of the sites with which we have to deal. The hills round the eastern end of Lake Trichonis are thickly covered with the remains of ancient towns ; in the narrow interval between the kastro of Palaiochori and the river Phidharis we find ruins at no fewer than eight points. We shall describe them in order, and then attempt to discover their connexion with the narrative of Polybios. / SITES ON THE EAST OF LAKE TRICHONIS. Along the eastern foot of the hill of Palaiochori a stream called the Mega Revma runs towards the lake from the hills in the south-western corner of the plain. High on their slope is the upper village of Makrymi ; at their foot is the lower hamlet, Kato Makrymi^. The eastern part of the lower village bears the name Helleniko, and the frequent discovery of architectural fragments, such as pieces of columns and capitals, and of Byzantine coins, proves that something stood here in the Middle Ages. A few blocks which contain sinkings for cramps might possibly date from an earlier period ; in the neighbourhood of the upper village ^ MaKpvvav," hva and Koto). Meletios, the good Bishop of Arta, who is responsible for much of the topographical rubbish embodied in modern Greek place-names, has the following in an enumeration of ancient towns in Aetolia [Geogr. ii. 306) : Uepavna n-dXir, ofioias ncu aSrai" H'Kevpavrj, "ApaxSos, MaKpiveia, Snov elvai rapa t& Xapiov MaKpvvov, Koi to \oara. He must have had in his mind chiefly Ptolem. iii. 14. 13. 240 CENTRAL AETOLIA. [ch. similar blocks are reported. As there are no traces of a fortified enceinte, we are right in concluding that this is not the site of an Hellenic polis : it may be that of some outlying village round a sacred place to which there succeeded a Byzantine church. The modern road to Naupaktos, five hours distant, runs from the south-eastern corner of the plain ^- Eastwards the plain is bounded by hills of limestone, the continuation of the Aetolian Alps, of which the last southern member is Mount Rhigani. Among these hills is the village of Kaliidhi ^, from which a deep gorge and stream run towards the north on their way to the lake. Just at the point at which this stream meets another coming from the south-east, we find two ancient sites. The first is on a steep rocky hillock placed in the gorge itself, at the confluence of the streams. The sides of the hill are so abrupt that on three flanks it has never been artificially defended. Traces of a single wall are found at the north-eastern corner, from which it con- tinues along the eastern side in a broken line, interrupted where the precipices provided sufficient defence. At the south-eastern corner there is a square projection, possibly indicating the point of ascent and entrance. The small circumference of the hill, and its excessively rugged char- acter, cause us to wonder of what use it could have been as a fortified place. It is also completely commanded by the towering sides of the ravine. The style of the work is careful ' irregular Hellenic ' of good appearance ; at the northern end of the line the masonry is fine ' regular ' work. Bazin^ is quite at fault when he describes these walls as displaying an 'art tout primitif, qui allait a la solidite sans etre capable d'y ajouter l'6legance.' The peasants know the ruins under the name KacrrpaKi, or Kao-reXaKi, the ' Little Fortress.' The site called Mesoviini* lies high on the ridge on the ^ The road ascends from the plain in a series of zig-zags along the sandstone slopes bounding the southern side of the plain. On the crest is the khan of Koup/ucKij. About one hour to the south is Mc7-a|a. Some future traveller must investigate the report of the existence of consider- able ruins at MetaxS. A variety of circumstances made it impossible for me to do so. ' KaXouSiov. ' Me'm, p. 323. ' Mea-ofiovvi. The enclosure seems to resemble that of the kastro xvi] MOROSKLAVON. 241 northern side of the ravine. The remains are those of a scarcely distinguishable enclosure among the patches of grain on the height. Looking across the gorge we see Kaliidhi to the south, Kastraki to the south-west, and the village of Dhervekista^ to the south-east, on the opposite hill-side. In order to reach the remains known as the Kistro of Dhervekista we descend from Mesovuni to the. ravine which joins that of Kaliidhi ; then we go eastwards until we emerge upon the small plain, called the KtJjuTros of Avarikos, at the other end of which gleams the white bed of the Phidharis. The small pjn-amidal height at the western end of the plain bears remains, and is called by the usual name, Palaiokastron. A tributary of the Phidharis skirts the base of the hill and turns several mills. The shade of the trees and the sound of rushing water make the spot a pleasant one ; it bears the appropriate name of Mawa Nepov, ' Mother of water.' The fortifications on the hill belong to the Middle Ages, but a few pieces of Hellenic wall, with square towers, prove the site to be that of a polis. At the northern foot of the eminence, close by the path, lines of beautifully dressed and cramped blocks have recently been uncovered. They may belong to a sepulchral memorial. Before we reach the kastro, at a distance of one hour from Mesovuni, we find a spring (Kephalovrysis), near which, under the deep shade of the trees, we discover about forty feet of excellent Hel- lenic masonrj', standing four feet high, and running north and south. Several ancient blocks lie round the spring itself. We may compare the similar remains near the source at the foot of the Berikos kastron ^ If we retrace our steps and follow the stream from Kastraki in the direction of the lake we at length again emerge upon the plain that lies at the foot of Palaiochori. A steep round detached hill rises on the edge of the plain, just at the mouth of the ravine; it bears north-west from Kastraki. Its level summit was surrounded in ancient times by a ring-wall built in rude style, of slabs rather than blocks. In the north-east of the enceinte, opposite the upper village of Morosklavon ; but it is hopelessly ruined. Apparently it has been in occupation in post-Hellenic times. ' Aep/S/xicrTa, AepiSe/ciTira. ' See p. 204. R 242 CENTRAL AETOLIA. [cH, of Morosklavon', which Hes on the hills above, are traces of a gateway of a simple but effective kind. It is formed merely by the overlapping of one end of the wall over the other, so as to create a short passage between them. The wall was apparently destitute of flank-defences ; but as the entire summit of the hill is now cultivated and the enceinte quite destroyed it is very difficult to follow the little that is still visible. A similar ring-wall enclosed a smaller hill lying immediately to the north, i.e. in the direction of the lake. Aetolian coins are picked up frequently on the double site. Although we speak of it as double, there is no reason to imagine the existence of more than one town here ; the hills are too small and too close together to have served as separate citadels of two distinct settlements, whereas if the one eminence was occupied as an akropolis it was absolutely necessary also to fortify the other. It is possible that the two enclosures were united by walls running across the depression ^. At the northern foot of the hill is the lower village of Morosklavon. Five minutes from the village, in the direc- tion of the lake, we find a Roman ruin among the vines. It stands six or seven feet high and is built of narrow bricks presenting a very beautiful appearance externally. Like most ruins of the same kind, it is popularly known as the Vay6ni, — ^ayevi, a 'tun or wine-press.' It is chiefly note-, worthy as resting on two courses of fine Hellenic blocks transported from some large building, very possibly from Makrynii. The modern high road passing through Lower Mor6sklavon does not climb the steep to the upper village of that name, but, turning towards the north-east, ascends the hills above the lake in a series of windings. When it has gained sufficient elevation it runs straight in a gradual ascent towards the village of Petrochori. The hill-side in which the road is cut is exceedingly steep, falling below almost sheer down into the water, and rising above no less abruptly. Landslips and winter torrents make continual repairs neces- sary. From Lower Morosklavon it takes one hour in order to ' Map6iTKKa^ov. ^ As in the case of Gyphtokastron and Petrovfini. XVI] KASTRO AND PLAIN OF PETROCHORI. 243 reach the kastro of Petroch6ri^ crowning a small height a short distance north of the modern village of that name. The houses of the village are placed on a knoll or swell on the very brow of the plateau above the lake. Towards the north-east a well-cultivated level valley extends nearly four miles to the heights at the foot of which Kephalovryson lies^. The plain measures about a mile across at its widest part, but its breadth gradually diminishes as we approach Kephalo- vryson. Along the south-east it is bounded by an almost continuous ridge, which descends until it stops opposite Kephalovryson in a low rocky tongue, behind which there runs a second narrow valley southwards in the direction of Chrysovitsa and Avarikos. A similar pass leads north- eastwards to Koniska in Kravari, a journey of five hours. On the northern side of the plain the hills are not con- tinuous, being interrupted by a depression between the kastro hill of Petrochori and the heights above Kephalo- vryson. The modern road from Kephalovryson goes through this depression northwards round the lake by way of Mokista and Guritsa, and so ultimately to Vrachori *- The kastro hill of Petrochori is under cultivation, and the terrace-walls in the cornfields are chiefly composed of the ruins of the old fortifications. The hill has also been occupied in post-Hellenic times, so that only very slight remains of the Aetolian enceinte are now distinguishable. The enclosure seems to have descended from the narrow akropolis summit on the north-west, and to have embraced the eastern and south-eastern faces of the hill. The area defended is small ; the circumference of the lines must perhaps scarcely have exceeded a third of a mile. The only considerable fragment now to be found of the ancient enceinte consists of the foundations of a large square tower on the northern slope, just at the angle formed where the wall turns up to the rock below the peak. The style in which it is built is good 'irregular Hellenic' The chief interest of the kastro lies in its grand situation. Perched hi_gh above the eastern end of the greater lake we look down ^ lleTpoxaipiov. ^ Ke(})a\6PpviTop. A pleasant village of over 600 inhabitants. It is the capital of the modern Deme IlafK^eia. See also p. 203. * It is a six hours' ride from Kephalovryson to Vrachori. R 2 244 .CENTRAL AETOLIA. [ch. the whole length of the central Aetolian basin between Mount Vi6na and the long line of the Zygos to the Akar- nanian summits low in the dim distance of the west. In the south-east we see the head of Mount Rhigani. In the east the view is limited by the rugged masses of Kravari. On the narrow plain below us, and almost due east, lies the kastro of Kephalovryson, which goes by the name of Palaiobazaro ^ (Palaio-bazari), or simply Helleniko. It is distant a short half-mile south-west of Kephalovryson, and about two miles from Petrochori. It is remarkable as dif- fering from almost all Aetolian cities in that it is built not upon the summit or side of a hill but entirely in the plain, at the foot of the ridge bounding the valley along its south- eastern side. In other respects also this kastro holds a unique position among the Aetolian sites. Its ground plan is that of a nearly regular quadrangle, three sides of which remain; the fourth side, that on the east, or more correctly the south-east, leans against the stony ridge, and perhaps never was defended, — at any rate the wall on this side, if it ever existed, has disappeared without leaving a trace. The other three sides, facing respectively north- east, north-west, and south-west, still stand to an average height of three courses. The thick growth of bushes along the top partially hides the ruins from sight, so that they do not exhibit an appearance in keeping with their real import- ance ^. At intervals of about one hundred and forty feet we find square towers, with a frontage of, in general, twenty- three feet. Fifteen such towers may be countfed. Near the western angle of the enclosure there occurs a semicircular tower, the only one on the site ; it is close to a square projection placed at the angle itself. Between this round tower and the square one at the angle we find a small gateway, of which there remains only the opening through the wall. The thickness of the wall varies between eight and nine feet. It is built in a style resembling that of Chalkis, i. e. the courses are, in general, regular, but the face of the blocks is left rough, and the joints are oblique. The blocks have a tendency to be nearly square; but in some ' To XlaKawira^apov, to naXato-/Jirafapi. To '"EXKjjvikov. ^ I regret that it proved impossible to obtain a satisfactory photograph, either of the site as a whole, or of any part of it. xvi] ANTIQUITIES AT KEPHALOVRYSON. 245 parts they are large parallelograms, as along the south- western side of the enclosure. The system of bonding the two faces by means of cross-pieces has been very carefully followed, so that the wall presents the appearance of a series of compartments, into which earth and small stones were tightly rammed. The maize-fields on the site, and still more their accom- panying network of small canals for purposes of irrigation, make it difficult to follow the traces of buildings within the walls. We notice, in the first place, that a road seems to have entered the enclosure at the eastern angle, where the line of the enceinte ends against the hill. This road probably passed quite through the site, following the base of the ridge ; its course is for some distance very plainly indicated by a series of rectangular sinkings made in the rocks by the side of the path for the reception of stelai. A long row of these cuttings stretches inwards from the wall as far as the remains now scarcely distinguishable as those of a church, that of the Holy Trinity. This church has been built entirely of ancient materials ; it is now only a heap of beautifully prepared stones, which have sinkings for iron cramps, and other indications of their having ori- ginally belonged to a building of superior style. The most careful search fails to reveal any certain vestige of the stelai that once lined the roadway. It is possible, however, that the mutilated block lying by the watercourse a few yards lower down may have been a member of the series. Lolling, in 1876^, deciphered one of the two inscriptions which it contains. It reads as follows: — Ilo\v o m >• S o O < A aKetas 8ie\5di/Tfr to areva, (Tvvip,i^av rdii MaKfboa-iv. 256 CENTRAL AETOLIA. [ch. Such an idea is likewise quite irreconcilable with the reluctance of the Aetolians to attack the king while he was ' on the high ground ^' and their desperate attempt to annihi- late the rear-guard as it descended into the defile. The Illyrians and peltasts were concealed 'under a certain hill on the descent ^.' The broken country on the crest of the plateau above the lake, between Petrochori and Morosklavon, offers abundant facilities for such an ambuscade ^. Most im- portant are the words relating to the ' descent.' We are at once reminded of the expression 'the whole ascent being thirty stades*,' which occurred in the description of the advance, being there wrongly inserted after the mention of Pamphia. The two are clearly one and the same ; namely, the steep and once perilous path up the hill from Lower Morosklavon to Petrochori. That is the only portion in the entire circuit of the lakes that admits of being described in a plain and straightforward sense as an ascent. The road from Petrochori onwards to Guritsa, on the other hand, crosses the plateau at a nearly uniform level above the lake ; at Guritsa we begin the descent towards Sobonikos. The conclusion to which we come is inevitable. Polybios has correctly described the narrow path along the forest-clad shores of the lake ; the situation of Metapa at the entrance of the dangerous pass, sixty stades from Thermon ; the position of Thermon itself, on the plateau, some thirty stades from Pamphia. By implication he has also given us the additional correct information that the defile came to an end at Pamphia, thirty stades from Metapa, and that at Pamphia the army swung round upon Thermon. The. sole difficulty is that ihis picture of the steep ascent, with 'the precipices above and below it, is inserted at the wrong place, viz. between Pamphia and Thermon, instead of between Metapa and Pamphia. * Pol. V. 13 : ev eva>vv)uov i7riaivov(rr]s. ^ Id.v.^: ' He himself collected the guides, and made careful inquiries of them about the country and neighbouring towns.' ' Cf Pol. V. 14. After reaching Limnaia, Philip ' proceeded to make a thank-offering to the gods for the successful issue of his undertaking. . . . His view was that he had ventured upon a dangerous country, and such as no one had ever before ventured to enter with an army.' * It is quite clear that all concerned were aware of the object of the inroad. 270 THERMON. [ch. donians arrived at Thermon later than two o'clock in the afternoon. Yet that is not the ' advanced hour ' of Polybios ^ Interpreting that expression without any preconceptions, we could not take it to indicate anything earlier than five o'clock, or thereabouts. Though he retards the army to the fullest possible extent, Leake thus finds himself in opposition to the historian as to the hour of arrival at the capital ; and, in addition, his position is entirely at variance with that of Polybios. The ancient account amounts to this : — the way was long arid difficult, and this, along with incidental stoppages, compelled the army to march at its best pace ; yet in spite of its exertions it was late in reaching Thermon. On Leake's hypothesis what is said is this : — the way was in reality short, but the army was so much hindered on the march that it was two o'clock in the afternoon before the destination was reached. Reduced thus to their simplest form, the two accounts are obviously quite incompatible ^. The return march furnishes us with additional objections to the hypothesis which we are criticizing. , On the first day of retreat Philip withdrew from Thermon to Metapa ; on the second to Akrai. On the third he moved to Konope. It was only on the fifth day that he marched up the Acheloos, and finally quitted Aetolia by the ford in the neighbourhood of Stratos. Now Leake, though he expresses himself with some reserve, puts Akrai at the point at which the Kyathos issues from the lake of Anghelokastron. We fail, however, to understand why the Macedonians, after their previous short stage from Vlochos to Metapa (the mouth of the Ere- mitsas, according to Leake), should have spent a whole day in passing over the three miles between Metapa and the point indicated as Akrai, and yet another day in traversing the three miles between Akrai and Konope. Then, again, ^ Pol. V. 8 : jjKe TToXX^s &pas irn rhv &epfiov. Note that the expedition took place in the latter part of summer (probably July or August). But the arrival at Thermon must have been calculated for at least an hour, probably two hours, before sunset, i.e. the disappearance of the sun below the horizon, not the commencement of darkness. In Greece, of course, the period of twilight is of very short duration. ^ Or we may put it in this way. Polybios implies that the march was a record one : the rate could not be beaten. Leake implies that the army could have done much better, or at any rate that it ought to have done so. xvin] BECKER'S TOPOGRAPHICAL SCHEME. 271 what was Philip's idea in visiting Konope at all? We do not read of any operations having been undertaken in its vicinity, and the movement was certainly not made on account of the ford, because the army subsequently marched northwards in order to cross the river. This being so, how was it that Philip did not take the direct route from Metapa to the Acheloos ? The advance from Metapa to Akrai and thence to Konope, and finally to the ford, is, on Leake's hypothesis, a detour over which three entire marching days were consumed, whereas the direct road from the mouth of the Eremitsas to the Aspro measures only from seven to ten miles. Not a word is found in Polybios to explain these difficulties, nor does Leake endeavour to supplement his account. Are we not justified in maintaining that for Poly- bios they did not exist, because he had no such route in his mind as that suggested by the English topographer ? Finally, we must draw attention to the fact that no vestige of an ancient town is to be seen, either at the point at which Leake places Akrai, or in the loW ground near the mouth of the Eremitsas, where he supposes Metapa to have stood. For Pamphia no site is even suggested ; but it is beyond question that there are no ruins to correspond to it in the plain below Vlochos, where alone they could be sought on Leake's view of the topography. We must remember that to Leake himself this could not appeal as an objection., It is ninety years since he penetrated Aetolia ; he could not tell what the future might reveal in the way of remains in the parts which he had not explored. We, however, know with certainty that no traces of towns are found where Leake imagined that they might exist. We must direct our search elsewhere. It is time to turn to the examination of a second hypothesis, that of Becker ^- Becker's idea is that Philip marched along the southern shore of the lake of Anghelokastron ; that then, turning between the two lakes, he traversed the causeway of Ali-bey, or rather its ancient representative, and so ascended to Vlochos. We have already thrown out the conjecture that in ancient times the appearance of the central basin of ' Becker, Diss. i. 27 fol. Aetolia differed somewhat from that with which we are now familiar. As we nowhere find in an ancient author the distinct and positive statement that the basin contained two separate lakes, it seems within the bounds of possibihty that originally but one sheet of water existed in the plain. If this were found to have been the case in historical times, Becker's hypothesis would be disproved at the outset. As respects his distribution of the towns on the route, Becker is as unsatisfactory as Leake. He puts Metapa^ either not far from Pappadhatais, or at some point more to the west along the southern shore of the lake of Anghel6- kastron. With regard to Konope, Lysimacheia, and Tri- chonion, he agrees with the results established by Leake. He also places Phytaion at Kiivelos (Paravola), and, like Leake, he altogether omits the name of Pamphia. In the case of Akrai he has a theory of his own : he puts it at Vrachori. He supposes the Macedonians to have retreated to Metapa across the land lying between the lakes, and next day, because they had already wasted the country between that town and Konope, to have made a retrograde movement to the northern side of the lake, in order to ravage the plain belonging to Akrai ^. Some of the objections to be urged are identical with those already brought against the theory of Leake. Becker's theory, like Leake's, does not give its natural sense to the statement concerning the position of Trichonion and Phy- taion on the right of the army. We hold it to be incontro- vertible that those two towns had been left far in the rear by the time that Metapa was reached ; whereas, according to \ ^ Becker, Diss. i. 30 : ' quam haud procul ab eo loco abfuisse puto, quo nunc Papadates vicus est, vel paulo infra potius in ipsa Hydrae lacus ora australi, ita ut fauces triginta stadia longae inter ipsam pontemque inter- cederent.' Becker gives the name Hydra to the lake of Anghelokastron : the ' pons ' is the causeway. ^ Ibt'd. 'Macedonas tunc demum fertilissimum campum, qui in laeva Hydrae lacus parte est, devastasse, veri est simillimum ; neque impedit quidquam, quo minus eos per fauces easdem regresses ipsum lacum circumiisse putemus, quum alterum tertiumque diem in hoc opere con- sumerent. In hoc igitur campo et, quod nomen urbis ipsum docet, in edito saxosoque loco Acras fuisse conditas puto, neque alium video ad haec omnia magis convenire, quam eum, quo nunc Vrachori urbs sita est.' xvin] BECKER'S THEORY CRITICIZED. 273 Becker's disposition of the names, not only Trichonion, as well as Phytaion, but Lysimacheia also, occurs after Metapa. Again, our criticism on the position which Leake assigns to Akrai and on the course which he imagines the invaders to have taken in their retreat apphes with even greater force to Becker's hypothesis. Not a syllable appears in the text of Polybios to suggest the strange vacillation of which, upon Becker's hypothesis, the foremost general in Greece was guilty. We are expressly told that a garrison was thrown into Metapa in order to ensure the safe passage of the army through the defile in both directions. What then was the object of returning to Metapa on the first day of the retreat and of combining with its garrison, only to march next day for the third time through the pass, thus facing once more the possibility of disaster at the hands of the enraged militia of Alexander the Trichonian ? Not content with this risk, Philip spends a whole day between Akrai and Konope, that is, according to Becker, between Vrachori and Anghelokastron, and another day between Konope and the ford by which he crossed into Akarnania. Could anything be more aimless than this marching and countermarching over the plain on the part of the plunder-laden army ? It is, in addition, quite irreconcileable with the historian's distinct assertion that the route of the return was the same as that of the advance ^. A spe'cial objection to the theory arises out of the question of the lake. Becker has correctly perceived that the route must lie to the south of the lake, but he has overlooked the fact that, according to Polybios, Metapa stood ' on the borders of Lake Trichonis ^.' Inevitably, therefore, if Metapa is to be located as Becker supposes, the lake of Anghelokastron must be identified with Lake Trichonis. The same result follows from the statement about the lake being on the left of the army. It was certainly the Trichonian lake that stood in this relation to the troops ; yet, on Becker's hypothesis, no other lake than that of Anghelokastron is on the left of the route. Nevertheless, he admits that Trichonion itself stood at Gavalii, which is a long way to the east of his route, on the borders of the larger lake, and out of all relation to that of ^ Pol. V. 13 : TTOioijicvos Trjv avTrjV iiravobov 17 Kcil irapeyivero. ^ Id. V. 7 • *'"■' oir^s Trjs Tpix<»vi8o! Xi^i -o < ffl H N < < < Id HI S O O h-l H U To face p. 285.' xviii] INSCRIPTIONS FROM KEPHALOVRYSON. 285 Asklepios, Dionysos, or other deity. Nevertheless, this record gains its value just because it is not according to the customary mode ; for the method of which it is an example is easily paralleled. It was the usual one at Dodona, for instance, where the formula is: — Sewa At^irjTi (or a^r\Ki) ikevOepov rov Mva, but the place of the ceremony is not given ^ In the inscriptions from Elateia, Daulis, and Koroneia, the formula is : — fv hvoiia eKKXrja-ia air)Ti., or h fwoixa e/cicXij(ria t&v avvehpuiv^. The inscription before us is doubtless one of this class, — a solemn engagement before the assembled people or their official representatives. If so, the conclusion that the site is that of Thermon can hardly be avoided ; for the stone would be erected at the place of the ceremony. It should also be. noticed that the upper part of the block contained an illegible inscription, of which Lolling deciphered perhaps enough to show that it was a decree of the League ^. Finally, we have the inscription published for the first time by myself*. It is a dedication made by the town of Opous and the Opountian Lokrians in gratitude to Lykopos, Strategos of the League. The date we do not know, but it may hereafter be fixed by evidence from Delphi®. The ^ Cf. Carapanos, Dodone ei ses ruines, Plates xxx, 1-5 ; xxxi, 1-4 ; xxxii, I. " Bechtel {Sammlung-Collitz), No. 1532 = Curt. Anec. Delph. No. 39. ° Mitth. d. arch. Inst. iv. 222. He made out ' wohl der Uberrest von kol\vov Ai[t;. 296 NORTH AETOLIA. [ch. Other sides it runs out intb lower-lying irregular spurs to which the name Sicoma is given. That below the above- mentioned fragment of wall seems to bear traces of having once been included within the fortifications of the main height. Numerous tile-fragments and terrace-walls on the northern and eastern declivities show that the town lay on those sides, but no inscriptions are reported to have been found. The kastro of Viilpi is the complement of that of Palaiokatunon. It bars the way to an enemy advancing from^ the north down the vale of the stream that falls into the Aspro near the bridge of Tatarna. This stream has no peculiar name. As is usually the case, it takes that of the village, being called the Vulpiotis. Velaora^ is reached after a three hours' journey to the west from Viilpi, by a rough path through the tangled brushwood covering the sandstone hills. The inhabitants of Velaora are dispersed in three hamlets ^, and in like manner there were three settlements in the plain in ancient times. Two of them are near the principal hamlet *. A rough crag with a narrow summit rises from the plain, about twenty minutes to the south of the village. A wall on the east and south-east cuts off access to the summit; the natural precipices, with the simple addition of a few blocks piled in the fissures of the rock, formed an adequate defence on the other sides. As it now stands, the wall is five courses high and about nine feet in breadth. Its. style is rude in the extreme ; if the stones had been larger it might have passed for an example of the so-called ' Pelasgic,' but as it is we can only regard it as a degraded form of that transitional style which is seen at its best in the kastro of Djuka. In this kastro of Velaora no attempt has been made to flank the wall by building towers or other projection: the line is nearly straight. The rough ground in front, together with the great thickness of the rampart, was deemed sufficient to protect the garrison. At the same time we must make allowance for the fact that the stronghold was probably only intended to provide a refuge in case of sudden attack; it ^ BcXacopa. In the common pronunciation sometimes BeKaopa ; whence the Velaghora of Leake, N. G. iv. 252. So in the French Map the name appears as Velavra, apparently a misreading or a misprint. " MaxaKahms. ' Machal^S row Taf^. a H D O X H s o 0. .4 > •I xix] VELAORA. 297 was not a permanent post, nor the centre of a polls. We draw this conclusion from the extreme smallness of the area enclosed, and from the absence of any trace of attempt to fit it for human habitation. Marked as is this latter feature in so many Greek palaiakastra, it is nowhere so noticeable as here : it is clear that we have to do with a place adapted and intended only for temporary occupation. The scanty remains of the second fort are found on an eminence, or group of eminences, five minutes due east of the hamlet. This en'ceinte was somewhat more extensive than the first, so far as we can judge from the scattered fragments of its foundations.. A tiny piece of wall reveals the style of masonry, and that also differs markedly from the masonry of the fort already described. It is in that quasi- polygonaP of which the ruins at Djuka furnish the best specimen. Probably we have in this ruin the main settle- ment, to supplement which the first-mentioned fort was built. The third palaiokastron of Velaora lies three quarters of an hour to the west, on a low ridge near the Acheloos. Its plan is a narrow quadrangle, the longest axis of which lies nearly north and south. The wall has fallen to the outside all round, and no part of it remains to any considerable height : its average breadth has been great, about ten feet. A cross-wall seems to have divided the enclosure into two parts. A gateway is placed in the western side, — a simple opening in the wall, which at that point exceeds the ordinary breadth. The masonry is of the type which we have called normal in this district ; it exhibits also the variations observed at Djuka and Viilpi ^. Topoliana^ is a village lying one hour and a half to the north-west of Velaora, across the torrent of Granitsa which cleaves a path through the plain westwards to the Acheloos. North of the village, a precipitous hill rises to a height of about seven hundred feet : round its summit we trace with ^ Bazin {M^m. p. 297), in describing the masonry of this second fort, says : ' II devait etre en cyclopeen regulier.' ' One or two courses ' still existed here and there when he visited the site. The fortifications are now practically obliterated. ^ In the case of this fort also Bazin's account is somewhat different. Its masonry is ' un beau modele de cyclopeen regulier,' but he remarked on the eastern wall a square tower ' d'appareil hellenique.' ' ToTToXioya, toiroKiava. 298 NORTH AETOLIA. . [ch. some difficulty the foundations of a small Hellenic kagtro. The position is a strong one, being defended by sheer cliffs on the south and west. Immediately below the walls a spring formerly existed, but it has recently been choked and destroyed by the outlaws. The chief feature of the site is the magnificent view that it commands. The palaiokastron of Topoliana is a fair specimen of those yet to be described in Northern Aetolia. Everywhere we find traces innumerable of the existence of a large population in this district in ancient times, but their settlements are known chiefly from their cemeteries. A few stones on a hill-top are the scarcely recognizable remains of their rude citadels. The relics of antiquity reduce themselves to a few miserable .stelai and ornaments from the tombs, not worth the trouble it costs to break down the barriers of suspicion within which the peasant possessing 'Apxaia entrenches himself The best of the finds are conveyed secretly to Athens or Patras, where they fall into the hands of the dealers. They are thus dispersed without any record of the place or circumstances of their finding; or, worse, they are falsely attributed to well-known or fashionable centres. Remains of a kastro are reported on a low height visible in the valley to the south-east below Granitsa: it is called 'Ay'ia 'EXeoSo-a, after the church erected upon it. Equally unimportant is the fortress of Lepiana ^, on a round hill forty minutes west of Granitsa, and due south of the conspicuous cone called Djuka which rises above the village of Lepiana. Tombs are discovered on the ridge near this kastro, in the place called 'PaKoj^Xau, and also at the foot of the village itself, in which a grave stele is preserved. It is a slab of sandstone, eighteen inches square, broken slightly at the upper right-hand corner: letters i" high. ANTIKPATEI/ 'AvTiKpdTei[a A N T I A I Ko Y 'AvtiS'lkov. Zelenitsa ^ lies three or four hours to the north of Lepian^, ^ Aeniava, ^ ZfXcw'ora, or ZeXewVo-a ; pronounced, like Lepiana, with tiie Albanian xix] ZELENITSA. 299 on the slopes of the Agrapha mountains, facing west behind the range of Granitsa (Mount Pteri). About twenty minutes to the south-east of the village there is a palaiokastron defended by sheer precipices on the south, and by deep ravines on the east and south-east. Access to the fortress can be gained only along the ridge to the north, and this path was cut off in ancient times by a strong cross-wall. A few fragments of terraces are all that we find on the site, for the enclosing walls have almost entirely disappeared. It is probable, indeed, that the artificial defences consisted mainly in the wall running across the neck of the hill, as it was almost impossible for a body of men to have approached the stronghold from any other quarter. The fortress is a striking example of the revolution caused in the principles of military engineering by the introduction of long-range weapons. Between the kastro and the village, which is distributed into several Machaladhais, there is g, large ancient cemetery, and many small objects of art are found in the tombs ex- posed by the torrents in winter. The whole of this region seems to possess considerable buried wealth. A.t Argyri, for example, a few hours to the west, beyond the river Platanias but east of the Aspro, bronze statuettes of some value have recently been discovered and sold to collectors'. At Zelenitsa itself a few iijscribed stelai have also been unearthed, but three only have been preserved. (i) On a slab of coarse grey sandstone, much worn, in fairly careful lettering. In possession of Ko'oras Za/cd/crjs. EYTYN EvTv^vofioi AINIXoY Aivtypv. KAEOTAS KXeoras TEA^MMoy TeXe/ifxov. guttural L. For the terminations -itsa, -ista (which are generally inter- changeable in modern Greek), see Leake, N. G. i. 283, noie i. ' From the descriptions furnished by those 'in the ring' in this neigh- bourhood, I have little doubt that the interesting Aphrodite belonging to M. Carapanos, and published by Lechat in Bull, de Corr. Hell. xv. p. 461, was actually discovered at Arg;^ri. The inquiries of Carapanos con- vinced him that its find-spot was in Epiros. A little money judiciously spent about Argyri would almost certainly secure results of considerable value. I got upon the track of several bronze statuettes, but time was precious and I had to abandon the quest. 30O NORTH AETOLIA. [ch. The first word was probably Eiravofios : The fragments of an o and an m appear on the stone. (2) Fragment of stele 15" wide, in possession of the Xpwa- oyeaipyaioi. Letters i" high, rather careless. Found' five years ago. TEAENlKoS TeXiviKos AroPAloY 'Ayopatov. (3) Stele, 18" wide, of a grey sandstone beautifully prepared for inscription. Letters very clearly cut. Found recently by the Chrysaphogeorgai'i, along with very fine gold earrings and other ornaments. EYPYNoMH Evpvvofir] T I M A r o P o Y Ti/iayopov. Two large blocks walled into a house in the 'farther village ' originally contained long inscriptions, but they were dressed with the chisel a short time before my visit, and not a letter can now be deciphered. The find-spot of these was not revealed. This disposes of the sites in the western section of North Aetolia. We might pass at once to the east, for between the river of Agrapha and the Megdhova we do not meet with any palaiokastron in the strict sense. The few sites that yield antiquities to the peasant excavator are not of any importance. Such are those of Marathias ^, Episkopi, and Stenoma. There is also a site between Kerasovon and Viniani, called Palaio-monasteri '■^ by the people, but the few stones still visible indicate a Greek building of some kind. At the village of Andhranova, two hours to the west of * Namely, the ' palai6kastron,' above the lower village (6 naTto fiaxaKas). Looking down the vale which opens upon the valley of the Megdhova we see a small plain, on the left bank of that river, at the foot of the heights occupied by the village of Kalesmenu. On the plain is the ruined monastery of Haghia Tri^s, near which ancient remains are also reported. ^ By the terms Palaio-monasteri (Palaiomonastero) and Palaioklesi {naXaioKKkfiai) the more intelligent peasants attempt to distinguish Mediaeval or Byzantine works, as opposed to Hellenic fortresses ('EXXi;- i/iKo, llaKainKatTrpn, *poupia). The test is two-fold, — the dimensions of the stones employed, and the presence or absence of mortar (TreVpaiy (xeyoXair, or TTfVpaif fUKpats /le aa-^earri). XIX] THE RIVER KAMPYLOS. 301 Prossos, antiquities are discovered, in a place called P6lemos. The Hegiimenos of the monastery at Prossos asserted that the finds included a bronze Eros and some painted vases ^ No great reliance, however, can be placed upon such descriptions; the lost or broken treasures of the villagers invariably far exceed in beauty and value the antiques that they can actually produce for inspection. The only interest attaching to this part of Aetolia is con- nected with the identification of the river Kampylos. We hear of it for the first and last time in Diodoros, who tells the story of Kassander's sudden appearance in Aetolia in 314 B.C. in order to support the Akarnanians. The object and result of this interference have already been discussed in connexion with the site of Agrinion *. Diodoros has preserved no information as to the locality of the encampment in which the conference with the Akar- nanian chiefs took place. The king came out of Macedonia, and did not at first descend into Akarnania. He encamped on the river Kampylos in Upper Aetolia ^. The reason for this move is not given by the historian, nor is it in itself very obvious. We should naturally imagine that it was dictated by the fact that Akarnania lay off the route which the king intended to follow; but this supposition is disproved when we find him advancing against Leukas, and moving home- wards by way of Apollonia and Illyricum. Now the geo- graphy satisfactorily accounts for Kassander's presence on the Kampylos. An obvious and natural route leads out of Macedonia through North Aetolia to Leukas. From Macedonia the army would first march into Thessaly, by the pass of Kalabaka and the Meteora monasteries, or by the more usual route through Tempe. The result was the same whichever road was taken ; in order to reach Leukas the Pindos had to be crossed, viz. from Trikkala or from Kard- hitsa. Kassander would thus strike the head waters of the M^gdhova, and following the course of the stream would ^ He showed me a small bronze figure of a dancing girl, clad in flow- ing robes, right breast bare, armlet on right arm above the elbow ; huge snake coiled round left arm. Poor work. ^ See p. 172. ' Diod. xix. 67. 3 : Aionep dva^ei^as ix MaKeSovias /jctoi Swdiieas /itydXijs ^Kev els AiTfflXi'ai', Kal Kana-rparimibevtTf irepl riv (caXou/ievov KaiiiriXov iroraiiov. 302 NORTH AETOLIA. [ch. fall at length into the trail which we have more than once mentioned as leading westwards from Karpenisi across the Megdhova, the Agraphiotikos, and the Aspro in succession. From the bridge of Tatarna on the latter river the road to Leukas crossed the Agrai's in a south-westerly direction. Knowing as we do from Diodoros that the Kampylos was in Aetolia, it follows that it must correspond either to the Megdhova or to the river of Agrapha. The former is the only stream really answering to the description implied in the name KajxTroXos, 'the tortuous^.' We must, therefore, identify the Megdhova as the ancient Kampylos. No justification can be found for applying the name to the Karpenisiotikos and that section of the Agalianos which flows between Kiitupas and Chelidhoni. Kassander had no need to descend so far to the south-east as to find himself on those streams. A point worth notice, as helping out our reasoning, is the fact that no fortresses are found along that section of the Megdhova which flows through Aetolia ^ This would render its neighbourhood peculiarly suitable for the somewhat protracted stay made by the Macedonians in the heart of the hostile country. We have thus found an answer to our question as to the whereabouts of the Kampylos, but the subsidiary problem awaits solution. Why did Kassander select this out-of-the- way region for his conference with the Akarnanians ? The surmise, that it was because he was not intending to advance with his main body into Akarnania itself, proves, on reference to his subsequent movements, to be unfounded. He could not reach Leukas without crossing North Akarnania. He must have pitched his camp on the Kampylos in order to paralyze Aetolian opposition while the concentration of the Akarnanian villagers was effected. The Aetohans could not venture to open a campaign in the west so long as the Macedonians threatened to swoop down upon the central plain from the north. Lykiskos, Kassander's heutenant, was thus enabled to carry out the Macedonian ^ Cf. Brandis, Mitth. uber Griech. i. 30: 'langs dem schlangelnden Flusse.' ^ In the upper part of its course the Megdhova flows through the country of the Dolopes, who would of course give the king right of way. They were always Macedonian in their sympathies. See Pol. xxi. 25. xix] ORACLE OF ODYSSEUS. 303 programme without let or hindrance. The moment Kas- sander removed the pressure three thousand Aetolians thirsting for vengeance gathered before the walls of Agri- nion ^ It would be interesting, and of some value, if we could by any means locate with certainty either the Oracle of Odysseus or the town of Oichalia. Our knowledge of the Oracle is derived from the Com- mentary of Tzetzes on the Kassandra of Lykophron ^. It appears that the yiavrfiov 'Ohvcra-im among the Eurytanes was alluded to by Aristotle in his Constitution of Ithaka, as well as by Nikander in his work on Aetolia. The Eurytanes, as we know them through Thucydides, are the last people in the world that we should credit with a knowledge of Odysseus. How can his name and cult have penetrated the mountains ojf Aetolia? We are here on the track of a connexion which we are unable to follow up owing to our ignorance of the true affinities of the races in Northern Greece. We know that the cult of Odysseus was established in Trampyia, a town of Lower Epiros*. According to one theory, firmly beUeved by Pausanias*, it was in Epiros that the hero at last appeased the wrath of Poseidon, among a people that knew not the sea nor the use of salt^. If the Eurytanes had in them a strain of Epirot blood they must have derived the cult of Odysseus also from that people ®. The Oracle of Odysseus is doubtless to be regarded as an example of the widely-spread dream^oracle ; and Odysseus, as the recipient of worship, must be akin to the under- ^ Diod. xix. 68. * Lykophr. Kass. I. 799: Mai/rty BevcKpov 'EipvTav crTei/rei Xems. Tzetzes says : 'ApiaroTeX^s tlujalv iv 'idaKrja-iav wokiTeia, Evpvravas iBvos tlvai Trjs Aitu- \las, ovofuurSiv diro Eipirov, trap' oTj elvai liaVTetov 'OSvaaeciis. T6 S" avTo Kai NiKavSpcJs (jyri(Tiv iv AItoKikois, ' Tzetzes in Lykophr. Kass. 800 : Tpafmrna, iroKis 'Hireipov, h6a nera v6aT0v '08viK€vvTas ay Kfiva ava