The Long Wall of Salona and the Ruined Cities of Pharia and Gelsa Di Lesina Author(S): R

The Long Wall of Salona and the Ruined Cities of Pharia and Gelsa di Lesina Author(s): R. F. Burton Source: The Journal of the Anthropological Institute of Great Britain and Ireland, Vol. 5 (1876), pp. 252-300 Published by: Royal Anthropological Institute of Great Britain and Ireland Stable URL: http://www.jstor.org/stable/2840891 . Accessed: 15/06/2014 22:28

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Royal Anthropological Institute of Great Britain and Ireland is collaborating with JSTOR to digitize, preserve and extend access to The Journal of the Anthropological Institute of Great Britain and Ireland.

http://www.jstor.org

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 252 Capt. R. F. BURTON.-On the From the AUTHOR.-IllustratedCatalogue of the Canterbury Museum. By JohnBrent, F.S.A. From Mrs. JACxSON.-Ethnology and Phrenology. Second edition. By the late J. W. Jackson.

The followingpaper was read by the author:

ThteLONG WALL of SALONA and the RUINED CITIES of PIIARIA andi GELSA Di LESINA. By Captain R. F. BURTON, H.M. Consul at Trieste. [With Plates xii. and xiii. and woodcuts.]

INTRODUCTION. ALLOWme to begin by expressingthe great satisfactionwith whichI findmyself once morein this room,and permittedto offeryou the results of three years' work. The specimenson the table will show you the nature of that work, and, before proceedingto the papers announcedfor this evening,I may, perhaps,be allowed brieflyto introducethem to you. The two sketches (P1. xiii. figs.1 and 2) representthe only flintimplements as yet foundin Dalmatia; and I have added the requisitedetails. They were shown to me by my learned friend,Prof. Glavinicof Spalato. I would proposehim and Dr. Lanza di Casalanza as correspondingmembers of our Society; and I will answerfor their value. The skull and the accompanyingbones were found near Bolliunz, a valley about fivemiles south-eastof Trieste. Here the Romans cut an aqueductin the live rock-a troughwhich supplied Tergestewith the best water. I have no doubt that someof the caves whichnow appear naturalwere hand-worked formortuary purposes; and, though I will not answer for the skull being Roman, or, indeed,of any great antiquity,I think that it may be a relic of the ancientrace, and, as such,I have brought it home for the collection of our learned associate, Dr. J. Barnard Davis. The collectionof potteryand the models of stoneimplements are intended for the admirable collection of our President, Col. A. Lane Fox. They are gatheredfrom the Castellieriof Istria, concerningwhich I latelypublished a paper,with many regretsfor the mode in whichit was published,and a heartfelt resolutionnot to do it again. The potteryis submittedto the judgment of experts. To an amateurit appears of different epochs,but I can answerfor the fact of its being an authentic find-most of it was dug up in my presence. It occursin the black earth.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Long Wiallof Salonat. 253

PART I.-TH-E LONG WALL OF SALONA. Salonawants but few words of introduction. She was in turn theRespublica Salonae (inscribed on leadentube of aqueduct); the Conventusand Coloniaof Pliny (iii. 22), and a hostof writers;the Roman metropolis of Dalmatia Felix, that fair and fertilesection of the land betweenthe Nestus or Tilurus (mod. Cettina)and the Naro (Narenta)rivers; a Proefectura et Pratura (FarlatiIllyria Sacra i. 27), with a Proepositus thesaurorum;a Procuratorginaeciorum; a Procuratorbaphi- orum(of thedyeing establishments); and a PraesesDalmatim (orFunctionarius perfectissimus). She was thegreat emporium of thecoast, the 6rtvetov, or naval arsenal(Strabo, vii. 5); and the "Totius DalmatimCaput" (Const.Porphyr. de Admin. Imperii, cap. xxx.) which, under Augustus,included a part of WesternPannonia. Virgil (Genethliacum,Ecl. iv.) sangthe birth of a " Saloninus,"and theglories of the paternal triumph:Horace (ii. 1) immortalisedthe honoursof lDalmatic or Delmatic victory. Here "Duke Bato" (A.D. 6, Dion Cassius,lib. lx., and Vell. Paterculus,lib. ii.) foughtfor libertyagainst the predatoryand oppressivemasters of the world; hereS. Domnius(S. Doimoor Dojmo)was sent, according to old tradition,by SaintPeter; and hereTitus, by order of SaintPaul (Tim.ii. 4-10) preachedthe gospel to Dalmatia. The remainsof what an Englishnovelist called the " smallbut prosperoustown of Salona" (?), thoughseldom visited, are, accordingto Prof.Steinbuchel, some of themost interesting of classicalruins. Finally, a highlyadvantageous position has made Spalato, its modern ournmtenens, the natural, whilst Zara is the artificialand politicalcapital, and the most flourishing, indeedthe onlyprogressive port of the old "iRegnodi Dal- mazia," whichstill formsthe southernmostprovince of the extendedAustrian empire. lButmy business at presentis witha singlesection of Salona, the" Long Wall," ofdoubtful and debatedorigin. The celebratedAbate Alberto Fortis (Viaggio in Dalmazia, e.c.2 vols.,Alvise Milocco, Venice, 1774, translated into English (London,J. Robson,1777), and French," Voyageen Dalmatie, Berne,chez la SocieteTypographique, 1778 "),' whotravelled in A.D. 1770-1772, and whosemeritorious labours have been a mine copiouslyquarried by laterwriters, has no notice of the "1iMurazzo,"or long,wall, althoughhe gives a detailed descriptionof theruins of Salona,in his vol ii. p. 42 (French translation,ii. sec.iii. p. 56).

* I shallgive in thesepages references to textaud Frenchtranslation for facility of reference. The book has becomesomewhat rare and costly.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 254 Capt. R. F. BURTON.-On the

On the otherhand, the late Mr. Paton, so well known as a traveller,and a writerof travels,and mentionedwith respect by the late Abate Carrara,visited Salona in 1846-7, and describes this most interestingfeature in the followingremarks (vol. i. 363, "Highlands and Islands of the Adriatic,"by A. A. Paton, 2 vols. London: Chapmanand Hall, 1849): "To the westward of Salona is a remnantof an immense construction,the origin and destinationof whichis quite unknownto the local antiquaries-a Cyclopeanrwall,* of regular quadrilateralstones, each from8 to 16 feetin length. At first sight, I imagined that it must have been the foundationof a temple; but, as it extends580 paces in length,I soon saw the fallacyof that opinion." The generic term, " Cyclopean," is given to the blocks 6 to 12 feet long by 2 to 5 high, at Tiryns, in Argolis, by Pausanias (KvcXrarcov. . . 6'pyovii. 25-7); but he also applies it to the hewn polygons of Mykenme,and even to the squaredmasonry of the Gate of the Lions. Euripides-repeat- edly adapts the same expressionto the walls of MIykeneand of Argos. Lactantius (vid. Stat. Theb. i. 252) explains it thus: " Arces Cyclopumautem, aut quas Cyclopesfecerunt, aut magni ac miri operis; nam quicquid magnitudinesua nobile est, Cyclopummanu diciturfabricatum." Of coursethe picturesque termwas seized upon by the imitativeRoman poets: Seneca, to mentionno other,says: Quid moror? Majus mihi, Bellum Mycenisrestat, ut Cyclopea Eversa manibussaxa nostrisconcidant. (Herc. Fur. iv. 996.) The firststep would be to smooththe joints, as in the artistic walls of Cosa, and the outer surface,as at Rusellke,after which the whole stonewould be hewn firstto the pseudo-isodom,and, lastly,to the isodomicform. Petit-Radel and Dennis (ii. 284) hold the polygonal structuresto be Pelasgic, and the former declares (Mem. Inst. iii. pp. 55-66) that theyhave been found as far southas Lucania and Apulia. Mr. Hamilton (" Arche- ologia") complicatesthe use of the wordby applyingit to four several formsof masonry,noting as manydifferent epochs. In the first,the gigantic,irregular blocks are of varioussizes, with smaller interstitialstones, but whollywithout mortar (Tiryns and a portionof the MAaltese " Torride' Giganti,"unhewn masses of coralline). The second shows masonry without courses, irregularpolygons whose sides fit closely (Mykenae,EBrurian

* The italicsin thispassage, and in the quotationsfrom Sir GardnerWilkinson and the Abate Carrara,are mine.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Long Wall of Salona. 25

Norba,Core, Arpino,Cosa, and Alatri; also Iulis (of Ceos) and Delphi; in the third,the strata are of the same height, but pseudo-isodomic,or unequal in lengthof stones(Boeotia, Argo- lis, and the Phocian cities); whilst, fourthlyand lastly, the blocks are of differentheights, but always rectangular(Attica). "Rectangular Cyclopean" sounds almost like an Irish bull. Perhaps we had better,with Dodwell (" Pelasgic Remains") and Sir William Gell (" Rome"), despiteBunsen (Ann. Inst., Rome, 1834), limit the term" Cyclopean" to masonrycomposed of irregularpolygons of large size, superimposedand fittedto- gether,more or less closely,with interstitial stones, but without mortaror cement. The oldest formwould be a massivewall formedby huge blocks of undressedrock simplypiled together without much care for jointing, as at Sidon, and in the Castellieriof Istria, especiallythat of Cunzi (Kunzi). Similar walls are found at Segni, Alatri,and other Etruscan cities,as opposedto those of Latium. In the second,the sides would be smoothedso as to correspond,but the stoneswould not be laid in truecourses; such are the ruins of Tiryns. The next travellerof note to Salona, the late Sir J. Gardner Wilkinson ("Dalmatia," &c.: London, lMfurray,1848), avoids using the terms" Cyclopean" and " Pelasgian," * and describes (i. 160-1) the " wall of large stones" in these words: "From this point" (the southerntown-enceinte) "another wall runs off,nearly at a right angle, which appears again by the roadside,about 115 paces to the westward,and, in a still moreperfect state thirtypaces furtheron. It is of verylarge stones, with bevellededges, admirably put together,and qf a stylethich resemblesGreek masonry. Some of the blocks are 13 feet long and 2 high. I traced it in the same direction to the distanceof 573 paces, or about 1,440 feet; and about 200 paces furtheron, is a line of rock resemblingmasonry, whichmay have been used as a continuationof the defencesof the city. On the northside of the wall, the sarcophagusof the Albucii familywas discovered,and other tombs are met with hereabouts. This wall may have been used to protect the entranceto the riverand the port,or may have belongedto the older city,before the Romans came into Dalmatia, when Salona was already a place of importance; and the characterof its masonryis the more remarkable,as it seems to point out a connectionwith the Greeks." The learnedauthor also showsthe importanceof Salona by quoting Strabo (lib. vii. 5, ? 5): E7retTia... . KatrcZ TcoVZIX/IadrE'aE v p7rapaXt[a,Kat To E7TLveLOV aTc0-v -dXAcv. "Eu-Tt8e& Tcov roXivvXPo'VO VoX6)s0a'vtcv 7rpos * Diorysius Hal. (i. 28) relates on the auithorityof Myrsilusthat the wa]l whichsurrounded the City of Athenswas called Pelasgic.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 266 Capt. R. F. BUJRTON.-On {Ate

PwoItalov9Tor MMo9 To'Vro, icaToucta9 8'E'OXcvaCtoX0yov9 EL? 7rEvTt/Kov'Ta, iv rtvas xa6 oXEt 6Xovas TE ICa\ ptva (Priomon?), Kal Ntvt'av,Ka\ XtvckzTtOV,TO Te VEOV Kal To 7raXato\v, as Eve,7rpqaOevo6 ZE/3ao-r (Augustus).* The highlydistinguished Egyptologist, to whoseliterary kind- ness I am personallyindebted, here showedhis normalacumen. But "'bevel" means properlyrabbeting, or oblique rebatement; in fact, "the angle formedby two surfacesof a solid body, meeting at an angle which is neither a rectanglenor half a rightangle." It is stilla favouritewith architects,but we must not apply the term to the masonryof Salona; the latter is chisel-dressedto a niarrowflat draught or border,and a boss or projection,apparently of unusualheight, ancd invariably forming a rectanglewith the lowerplane. By some writerson the Holy Land (e.g. Dr. Barclay, " City of the Great King," p. 494) it is made a characteristicof "Hebrew architecture,"whatever that may be; and, probably because they observed it at the "wailing-place" of the Jews, they named it the "Jewish bevel," a compoundmisnomer, it being neither" Jewish" nor a "bevel." The " Chosen People," I need hardlysay, borrowed all their architecture,and, indeed, art in general, fromthe polishedPhoenicians, and even moredistant races. This varietyof stone-dressing,so usefulin the determination of styleand date, appears hardlyto have been studiedwith the attention which it merits. A distinguishedEnglish archae- ologist at Rome informedme that he held it to be a classical

X~~~~~~~~~~~~~~~~AC1 II:9

formoriginating in the early centuriesafter our era. How far wronghe was, may be judged fromthe foregoingsketch of the huge walls, which are generally,indeed universally,supposed

* ItThen . . . the coast of the Dalmatn, and theirnaval arsenal,Salon. This nationwas for a long timeat war with the Romans" (alludi-ngto the Illyrian wars). " They had fiftyconsiderable settlernents, some of whichwere in the same rank of citiesas Salon, Priamon, Ninias, and the old and new Sinotium.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Long Wall of Salona. 257 to be partof the Agger of Servius Tullius. I was honouredwith the illustrationand the followingnote by the discoverer,the Right Rev. Father Joseph Mullooly, the learned Prior of S. Clemente,whose " Historyof Excavations," and " Saint Clement and his Basilica in Rome " (1 vol. 8vo) are so highlyvalued: " The breadthof the draughtor borderof the tufastones is a little more than an inch each way, and the cavityor channel between them precisely three centimetres,which gives the height of the boss. The tufa blocks,which I discoveredunder the floor of the subterraneanbasilica are believed by Mr. Parker, C.B., to date from the early days of the Roman Monarchy,and the travertinefrom the Republican times." The object of the "draught and boss" was evidently to relieve the jointing fromthe over-pressureof heavy bodies; hence we find the systemvariously adapted to a multitudeof differentforms by the latest,as well as by the earliest,builders. Some are exceedinglycomplicated. For instance,in the palace of King Hensius, at Bologna (Palazzo del Podest'a,A.D. 1261- 1486), each stone has a central rose upon a flattable, highly raised,and connectedwith the cornersby fourribs, each from one of the sides of the square. (See fig.) Perhaps the most popular is the bevel with rusticatedboss; and I have noticed,both in Eng- land and in Syria, the raised draught and the sunken centre, a modificationwhich clearly defeatsits own purpose. The learned Dr. Rosen, formerlyPrussian Consul at Jerusa- lem, findsthe followingthree distinct epochs of mnegalithicand "bevelled" (i.e. draughted and bossed) architecturein that mostvenerable part of the East: 1. The fine,rather wide, and shallow draught,often 6 inches in breadth around the edge, the whole stonecarefully squared, and the boldlyprojecting boss finelycut and plain-dressedlike the border. This he would call pre-Herodian,though there is no reasonto thinkthat the Israelitishcities had any peculiari- ties of architecture,and he instances the "Wailing Place;" Ar6kel Amlr,in Gilead (B.c. 290), and otherwell-known ruins. 2. The deeply-draughtededge, with the face of the stonepro- jecting moreboldly, and only rough-dressedwith the hammer, or leftas hewn fromthe quarry. He would considerthis early Roman and Herodian, and he finds it in most of the ante- crusadingremains. 3. The latest type is the roughly-draughtedborder, with the Augustusbiu'nt them clowni." In Strabo,the Dalmnatiancoast, the coast of thc Iapodes (Fiume), Liburnia (NorthernDalmatia), anid the LiburnianIslands, especiallythe modernLissa and Lesina, are givenin due order.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 258 Capt. R. F. BURTON.-On the

whole face of the stone standingboldly out, but only hammer- dressed,or leftas firstquarried. This, he says,denotes the later Roman period. The Rev. Mr. Tristram,a careless and superficialobserver, speaks (p. 78, "The Land of Israel," London, 1866) of the 'well-knownJewish or Phoenicicanbevel," and attributes(p. 80) to Dr. Rosen's firstand earliest epoch the Haram of Hebron, whichwe have everyreason to thinkis of far later date. I woulddistribute the draiighting and bossing(not bevelling) of the Syrianruins into three epochs-the Phoenician,*the classic (Greekor Roman), and the modern,the latterbeing conspicuous in the khansor caravanserais. There are manylocal varieties; forinstance, the double formin the churchof Constantineand Helena at Yabruid. Syrians,ancient and modern,work every- thing,from sewing to stone-dressing,in ways differingfrom, and oftencontrary to, Europeans. For instance,they begin to dress,not with the hammer,but with the pick, whichwith us comesmuch later. The modernstyle of cutting stone in Dalmatia and Istria,which probablydates fromthe remotestdays, may throwlight upon the systemof their classical predecessors. The ashlar is first dressedwith the heavy squlare-headedmazza or metal hammer. It is then subjectedto the puGnta,pointed, or narrowchisel, or to the finerscalpello, both used with the mnazziola,or nmaglio,a heavy mallet of soft iron,in which the hollow can be filledup. The modernpavement of Triesteand other townsemploys the punta when the sandstoneblocks are laid in place. The next processis the applicationof the brocca,a triangularpick, with sharp apex and a toothed base. Then comes the miiartello di dentigrossi, medii (vuilgo, " it bastardo") e fizi, thelatter called doppia mar-tellina,because the teethare double in number; for instance,one head will have sixteen,and the opposite eight. The Frenchboucharde with diamond-shapedsteel points,worked with the masse,and the Italian boeciarda,useful in treating graniteand hard stones,is commonat Vienna, but not used at Trieste or in Dalmatia. Finally, the stone,finished with the finerscalpello of many forms,even dentated,is polished with commonsand or saldame(Nolla or Sasso Marzo), a finepowder of silex foundin the limestonestrata, and not easilyaccounted for.t * From the Phoenicianscame the Etruscanboss, which is foundat the ruinsof Misalno,near Bologna, to mentionno otherplace. t These pockets,ful of finesilex, are mysteriousformations, to whichLinnseus (Syst. Nat. " Silex") alludes," Silex nasciturin montiumcretaceorum rimis, uti quartzumin rimissaxorum." We findthem at Proseco anclReppen Tabor,near Trieste,where the colouris dark,and near Pola, whose ruddy or straw-tinted sand has been extensivelyused forVenetian glass. The same anomalyoccurs in olcanic Iceland,whose silex beds near Cape Reykjanes(the south-westernex-

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions LonigWall of Salona. 259

If a spigolo(list), or a listello(draught), be required,it is worked withthe martellorather than the scalpello,and the rtsticais left simplyrandom-tooled or hammer-dressed, not grooved nor pitted into holes (prison rustic). La Bugna (7Jietrebugnazte) is the termapplied indifferently to " frostedstones," to the bevel,or to the boss and draught. Amongst Dalmatian writers,the only authoritywho has treatedthe " Murazzo" ofSalona as it deservesis thewell-known Abate Carrara (" Topografiae Scavi di Salona," del Dr. Fran- cesco Carrara,Trieste, 1850). In the first(topographical) part of his learned littlevolume, he mentionsit only once (p. 63); in the secondportion, or historyof the excavations,he refersto it three times. The first (p. 128) is in connectionwith the classical cemeteryenclosed by it, and by a wall of large white (limestone)blocks, carefully squared, running parallel with it to the north,9anddistant 3 Viennesetese (fathoms. each 6 Austrian feet-6 feet2 inches -67). In p. 147, he tells us that between 1847-8 were opened " ottocento klafter(the same fathom)di muro ciclopico di epoca antiromana;" and in pp. 136-9 he describesit in thesewords: " All 'ovest della citta antica si mostravanogli avanzi di un muroa grossimassi squadrati della quarta epoca delle costru- zioni ciclopiche(courses of various sizes, but alwaysrectangular). Il quale continuavainterrottamente per quasi 130 tese, senzza miostrarenm' cornihciarnento niewfie. Tra per la curiositadi deter- minarel'estesa e la direzionedi quel muro,non meno che la sua relazione col perimetrodell' antica citta,e con cio satisfaread un importantecuriosita nostra, ed al desideriodel dottoviaggia- tore Scozzese, A. A. Paton (' Allgemeine Zeitung,'Adriatische Briefevii. Salona, num.141, 1846); e trapel desideriodi cercare la lunghezza della necropoli discopertaa fiancodella via mag- giore, condussi a termine,mediante tasteggiamenti, uno scavo importante. Dal quale rilevo che il murociclopico seguato nel mio piano a linea grossa interrotta,lungo pic di 800 tese,par- tendodalla cinta a ponentedel perimetroantico, corre, pressoch' in linea parallela alla stradaregia, attraversandoil torrenteSlano, e continuandosino a quello di Blandiste,che marcail confinetra Salona e CastelSuguraz. Di la, anzicheavanzare al disopradella strada,risalta al Sud, riescendocon dolce deviazionedalla prima linea finoalla localitadetta Stagun ove terminacon due mausolei. 0 . . . La maisuramedia de' massi che formanoil muro ciclopicodA in altezza 2' 3" (Viennesefeet 100= 103 71 English), tremity),a purely eructive coulntry,have been spasmodicallyworked. Pliny (xxvi.6) declaresthat the Roman workmenused a sand found in the bed of the Adriatic when the waters retired. Fortis (ii. 271) noticed this Saldamo at Loparo (Neo-paros?) in the Island of Arbe.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 260 Capt. R. F. BURTON.--O) the inlarghezza 2' 6"; in lunghezza10'; il muronon e piugrosso della grossezzadei massi. " Codestomuro ciclopico fiancheggia la stradaantica romana soprala quale i francesi,per ordine de Marmont,aprirono nel 1808 la stradaregia attuale. La deviazioneche ho notatadi essomuro della stradaregia al torrenteBlandiste, si spiegacon ci6,che i francesi,arrivati a quel punto,piuttisto che progredire sullalinea della via antica,rifecero la direzionedella stradaper farlapiu mediterranealungo la bellissimariviera delle Castella. Cio nulladimeno,dalle osservazioni fatte in tantianni a Salona, vedocostantemente che ognistrada e viuzzaattuale corre sopra le rovinedi una via o di un clivoromano. " Dal murociclopico preesistente a Salona romana,trassero partitoi Signoridel mondoper istabilirloa lineadi divisione frala via pubblicae la maggiorenecropoli. E difatti,dal punto in cui il murosi staccadal recintodi Salona,sino a che siperde nel torrenteBlandiste, al norddi esso,trovasi il cimiteroantico, al sudla strada; da Blandistea Staqunla necropolisi presenta al sud,la stradaal nord. I1 quale mutamentodi disposizione risultanteda moltissimitasteggiamenti da me fatti(a tutto giugnodal recintoall' ovest di Salona sino ai mausoleiche seguanoil terminediscoperto del murociclopico si mostravano piiutdi 300 buchinon minori in superficiedi una tesa quadrata, profondidai 4 a 12 piedi),si spiegadi leggieridal contrasteche offronole due lineedi murotoccanti al torrenteBlandiste. Del restoper determinare l'uffizio diquel muro converrebbe continuare l'operaincominciata nel quarantotto,e per lo menocon ispessi tasteggiamenticercarne l'ulteriore andamento. S'arriverebbe forsea conoscerla,come e verosimile,quale murodi precinzione di Salonaantiromana, che e quantodire dell' anticaSalona." In p. 139 we read," Ora nel discoprimentode piu che 8000 (fathoms)del murociclopico, hassi un mezzodi comprenderela causa del fallo,e scusaretaluno di que' grossissimierrori" (alludingto thediscrepancies of shape and measurementsfound in formerdescriptions ofSalone). Page 157notes that the Roman theatrehas a " muraturaa bugnato,"which, I havesaid, applies equallyto thebevel and tothe draught and boss, the latter being here meant; but no noticeis takenof the same workin the Cyclopeanwall, which is farmore remarkable and characteristic. Moreover,when statingthat the "iMurazzo" showsneither beginningnor end, the learned Abate had notnoticed the highly interestingeastern prolongation which extends nearly to the newtown. Theaccompanying plan ofSalona (P1. xii.) showsby numerals theposition of the pre-Roman wall. It begins(H) at thejunction ofthe two enceintes, the " linea di congiungimentode' dueperl-

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Long Wall qf Salona. 261 metriRomano-Salonitani" (Carrara), and near where the learned Abbe places his " Porta Suburbia." The stones,at once dis- tinguishedby theirsize and by theirblue-grey tint, the argilla marinaplmbata of old authors,are muchdegraded. Thenceit runsfrom magnetic SSE. to NNW., roughlyspeaking: beyond theamphitheatre forming the north-western angle, it has been totallydestroyed; but farther along the path, the line bends at nearlya rightangle to the south-west,till it abutsupon the modernStrada Postale, or Regia of Spalato-Trau.*Hlere also thesoft material has been seriously inijured by time and weather. Beyond this point it resumesnearly its originaldirection (SSE.--NNW.), and, passing the place markedin the plan "1Scavi di, 1824,"Orto di 1VIetrodoro,it is in tolerablepreserva- tion. At theangle where the StradaPostale bends suddenly almostto north,a naturalreef-ridge of large limestoneblocks, standingup likea wall,prolongs the "iMurazzo," with a slightde- flection,tothe sea-shore. I saw notraces of work on thisfeature, and, althoughdraughted and bossed stonesare said to have been foundalong the easternside of the Frenchhighway, I couldnot detect them. The accompanyingrough sketch. from my diaryshows the positionand the dimensionsof the threecourses of masonry wherethe wall is bestpreserved. I was assistedin thework by M. AristideVigneau, of Spalato,and I have been promiseda photographof thishighly interesting section. Heightof highest tier . 0-61metres } 2 ft.English Heightof middle tier . 061 metres2 f Heightof lowesttier . 018 metres 7 inches The latterburied in gutterof roadside. Mostlyblackbelrry bushes.

2..72.'f Z.76 ' 1.5 zi -' -A.~~~~~~~~~~~~~~~~~~~~.

"Murazzo" (Long Wall) of Salona, facingSSW. The loweststratum is almostconcealed by thenorth-eastern side-drainof the StradaPostale. The reverseflank is com- pletelyburied, and the bushesprojecting over the top spring

* The well knownTraguriulm of the Romans,which Const. Porphyrogenitus (De Administr. "Imperii," cap.xxxvi.)writes Tetrgurium; we find "Cum tota Travia " in a documentdatilng from A.D. 1400 (Boglic,p. 91) ; themoclerln form is Troghir(Slav) and Trait (Ital.), whichWilkinson writes, without reason, Tralu.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 262 Capt. R. F. BURTON.-O)? the

froma cornfield. The 8oonerexcavations are here made the better. This " Mlurazzo,"to use the Venetian term,cannot be con- sidered a sea-wall,nor is it the defenceof the more modern Roman city-the two favouriteconjectures. The shoreline is wavy, and its distance from the masonryranges between a minimumof 50 and a maximumof 220 metres. The Strada Regia followsvery nearly the ancient line, as the discoveryof a milliarycolumn proved,and the followingfacts show that it was like the Via Appia and the Roman entranceto Palmyra,*a favouritesite for cremation and interment,ash urns beingfound on one side and tombson the other. North of the wall is the well-knownRoman necropolis,where a numberof sarcophagi, based eitherupon stonepavements or upon the naked earth,are still seen. They bad been rifledand injured,probably about the time of the Hijrah, by the Avaro-Slav invader (circa A.D. 639), beforehis expulsionby the Croat. On the otherside again (south),at theplace marked"Scavi, 1824,Dolj Sepulcrali," a curiousamphora-shaped dolium, with narrow neck, was found horizontallycut for the admission of a human body. It is figuredby Dr. Lanza (" MonumentiSalonitani inediti"), but I could not procurea copy,the book havinglong beenout of print. Here many sepulchralurns with ashes, and mostly inviolate, unlike the moreexposed sarcophagi,have been unburied. We may, therefore,conclude that this part was the Ustrinum, Ustrinator burning-ground,the Smashan of the modern Hindus. It is not a littlecurious that no less than ninejars containing bodies have been found when excavating the cities and ceme- teriesof Etruria Circumpadana,near Bologna. The Bolognese amphorcewhich containedthe remainswere eitherfull length -that is, pointedat the base-or half-sized,with flatbottoms, and in all cases the skull was foundupwards or near the mouth. The reverse was the system of the Brazilian Tupi-Guaranis, and I have suggested ("Notes to Hans Stade," i. 125) that the idea arosefrom their desire of returningthe bodyto theposition in whichit was originallyformed. In all cases whichcame under my noticethe head was placed downwards,as if in the womb,

* From Palmyrawe mayjudge that the practice,at once hygienicand aesthetic, passed overto Etruria,from which it was borrowedby the Romans. The Etrus- can cosmogony,which puzzled antiquariesby its curious resemblancewith that of Genesis,has, at length,been explainedby the discoveriesof Mr. GeorgeSmith at Eoyunjik. The Chaldean storyof the Creationand Fall of Man provesthat the Genesiticmythus attributed to Moses extendedthroughout the nearerEast fromEgypt to Assyria,and henceit was evidentlytransported by the Etruscans to Italy. t The Ustrinumdiffers from the r'ugov or Bustum; in theformer the corpse was burnt; in the latterit was also buried.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Long Wall of Salona. 263 althoughM. Adolfode Varnhagen(" iistoria Geral do Brazil," P1. opp. p. 112) gives an illustrationwith the feetdownwards. That the " Miurazzo"is notRoman is provedby its remarkable contrastwith the more modernepoch. The materialis every- wherea calcareouseocenic marl,a transitionfrom limestone to sandstone,blue-grey, and easily degraded, the marneof the Mons Caprarius,which formsthe loftybackground of Salona. All the Roman citv is built of the calcaire (nummulitic,*hip- puritic,&c.) of the same period,quarried fromthe backboneof Dalmatia, the eastern forkof the Apennines. This, in fact,is the normalrock bounding the Mediterraneansea-board. The material of Diocletian's palace was quarried in the island of Brazza, once doubtlessterra firma, and the port at whichit was embarkedis still knownto the people as Spliska, Spljet being the Slav equivalent for Spalato or Aspalathus.t Again, the lengthof the sandstoneblocks which we measured,varies from a miinimumof 0 90 (2 feet 11 inches) to a maximumof 2176 metres(8 feet2 inches); the depthof the highest and middle coursesis 0-61 (2 feet),and the thirdshows only 0418 (7 inches). The widthof the draughtranges between6 and 16 centimetres (2-40 inchesto 6A40inches), and the bold boss may have originally risen as high as 8 inches. In the Roman theatre(P) the stones are also draughtedand bossed,but the largest gave a lengthof 1111metres (3 feet7 inches),with a draughtvarying in width from 31 to 87 millimetres(1 inch to 4 inches the maximum),and the projection of the boss was insignificant. The same stone-dressingwill be foundin theTemple of Escula- pius at Spalato-a name traditionallygiven withouta shadow of reason,and in the " Duomo" (domts or cathedral)of S. Doime, attributedto Jupiter,apparently because Diocletian assumed the title of Jovius; or to Iiana, because the frieze showshunting and otherrural subjects; but in both cases the draughtis narrow,little exceeding an inch. Two dlistinctorigins have been proposed by local antiquaries forthis interestingfeature, and both agree,with Wilkinson,in attributingit to the Greeks. P. Farlati (Illyr. Sacr., i. 272) and Carrara (loc. cit.) findit * These nummulitesmark the deblutof the Tertiaryepoch, and the hippurites are so commonin Istria that theyhave been called the Istrianformation. t Const. Porphyro.gives Aspalathos (chap. xxxvi.) amongst the coast cities held by the D)almato-Romnans.In the fifthcenitury (temp. ArcadinLsNotit. diginitat.utriusqueL Imperii. caps. ix. x., quoted by Lanza, p. 23, Dell' Antico Palazzo, &c.), we read of the " ProcuratorGylnmcii Jovensis Dalmatine Aspalato." According to Lanza, the " urbicola," rebuilt after tbe Avar invasioniof the seventhcentury, was firstcalled Aspalathum,then Spalatum, and lastlySpalato, not Spalatro,as the learnedFortis has it. Mr. G. P. R. James,the novelist, speaks of the "little village of Aspalathus" in the days of Attila; he also terms Salola a "small butprosperous town "-the "but" is charming.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 264 Capt. R. F. BLTRTON.-On the in ApolloniusRhodius (B.c. 250). Thatpoet (iv. 336-563 etseq.) makes,in his "Argonautica,"the Colchi, led by Absyrtus,who accompanied,or who was senitby, his fatherRAtes, in pursuit of his sisterMedea, occupy the islandin whichhe was slain. Thisis generallysupposed to be Osero,or Ossero,near Cherso (Kherso)' in the Sinus Flanaticus(hod. Quarnero, or Gulfof Fiume,and its section between Cherso and Arbe the Quarnerolo). The earliername was "A*copos,, "A+opo9, or "Afoppos ; and the neighbouring"AfvpTevs, Ausyrtides, or Absyrtides(cf. Strabo vii. 5, vol. i. p. 484, Bohn) as faras the Nestus(Tilurus) river of Cettina,tnot to be confoundedwith the Nestusof Epirus (hod.IlMesta or Kara-suf),a coast wholly wanting islands. After occupying CC"usqtuead Salanigo)em(Salona?) fluviian, festidaque terramn"(Carrara), they moved off to Issa (Lissa) to Kerkyra Nigra (KeplcvpapeXatva, mod. Curzola) and Melite(Meleda). In lib. 41,1. 524 (Merkel'sEdit. Leipsig,1854) we findmention of 'TXX,qi9,and in line 535- "A,lt 7ro'tXv"Aryav'v 'T-Xqi&a " wheresome read J'yavqv,and others,ueey6Xqv," the greatest of the fifteencities of ScvmnusChius (407), and possibly hence the corruptionSalang,on, twice referredto by Carrara (pp. 1-4). In lines 562-3 we have 'AXX' eOeov(yafl 9 'TXXqfl'o&E4avtoVTCE TrX6Ot And this Hylleis is supposed to have been colonised by Hyllus, son of Hercules (B.C. 1230), fatherof the 'TXXe^t or CTXXeO The learned Abbe adds that if the Siculo-Issani (of modern Lissa) built,as we know they did, Epetium (mod. Stobrez) and Tragurium, afterwards the "oppiduin Roinanorunr nmarmore notun)' (hod. Tra'i), they would hardly have neglected the admirable position of Salona, wvhichlies betweenthe two, and

I The Abbe Fortis," Saggio d'Osservazionlesopra l'Isole di Chersoed Osero," Venice,1771 (pp. 1-12), treatsthis subject with abundant erudition. I fail to see the reason why a barque,manned by about 50 men,should not have coasted along the Black Sea, have ascended the Danube, anldeven have been portagedto the Istrian coast. Yet the " Mythof the Argonauts" is a favouritethesis for Germaninebulosity, andcl the last treatmenitwvas admilnisterecd by Dr. A. Kulhn (" Uber Eutwickelungsstufender Mythenbildung,"Berlina, 1874).' t "Nastus (&lii Nastos) urbs et fluviusIllyrii hilc Nestius," says Stoph. Byzant, quoted by Giovanni Lovrich, "Osservazioni sopra diver-sipezzi del Viaggio in Dalmnaziadel Sig. Ab. Alberto Fortis," &c., Venezia, 1776. He is a severecritic who choosesfor his mLotto," His quw narratasunt niondebemus cito credere; muLltiementiantur ut decipeant,multi quia deceptisunt " (Senl.de ira) ; and " Credat JudaeusApella " is a favouriteexclamation. The learned Abate's book was so famousthat it could not fail to findits Zoilus.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Long Wall qf Salona. 265 which is far superiorto both. Moreover,he assures us that Illina is the name still locally given to the ground lying west of the oldest Salonitan gateway,the Porta Cesarea; and that Illino-vrilo(Hyllus-fount) is the peasants' name for the spring betweenthe chapels of SS. Cajo and Doimo at the footof Mons Caprarius. On the otherhand, the Herakleia of Scylax Caryandensisis proposedby otherantiquaries, especially by ProfessorFrancesco Dr. Lanza di Casalanza,- a highly distinguished geologist,

* The followingis a detailed list of ProfessorLanza's meritoriousworks: 1. "In Cyanuretumrubium, inquisitiones chemico-pharmacologicae." Ticini Regii,1831. Vol. i. in 8vo. 2. " Sopra le Termedell' antica Salona." (V. Bullettinodell' Istituto di corres- pondenzaarcheologica. Roma, 1837,pag. 131). 3. " Relazione nosograficastatistica sull' epidemia colerosache invasela Dal- mazia nell' anno 1836, corredatadi osservazionipratiche specialie generali, aggiuntavila descrizionedell' Aretermoinventato dall' Autore,per la immediata applicazionedel caloreall' esterno."Trieste, 1838, un opuscin 8vo,con una tavola. 4. " Saggio storico-statisticomedico sopra l'antica Narona e lo stato presentc del suo territorio,corredato di una cartatopografica." Bologna, 1842, vol. i. in 8vo. 5. "Doveroso tributodi un figlio." Torino,1846, un op. in 8vo. 6. "Dell' Mausoleo dell' ImperatoreDiocleziano in Spalata: Lettera all' illustreSig. Cav. Geo. Dr. Labus, Spalato, 1 Agosto,1846. (V. Giorn. La Dalma- zia, 1846, n. 52. 7. " Ancora pochi cennisul MauLsoleodi Diocleziano." (V. Giornale,La Dalb mazia,1847, n. 14. 8. " Confutazionealla illustrazioneclel supposto sepolerodi Diocleziano e vera interpretazionedei bassirilieviche vi si trovanoscolpiti." (V. La Dalmazia, 1847, n. 48-49. 9. " Le B3occhedi Cattaro,descritte ed accompagnateda varicostumi." (V. " il Mondo illustrato,"di Torino,ed il Giorn.,La Dalmazia, del 1847.) 10. " I MLontenero."(V. Enciclopediapopolare di Torino.) 11. "Dell' Isola Lesina." (V. Encic. pop. di Torino.) 12. " Dell' Isola Lessa." (V. Encic. pop. di Torino.) 13. "Narenta." (V. Encic. pop. di Torino.) 14. "Dello stato economicodi Castilnuovonel Circolodi Cattaro." (V. Gior., La Dalmazia, 1847.) 15. "Sugli attualibisogni della Dalmazia: Letteraad un amico." (V. La Dal- mazia costituzionale,1848, li. 16.) 16. " Delle condizionipresenti dell' Austria: parolidi un dalmato." (Art. pub. nel Giornale La Dalmazia costituzionale e reprodottonel Giorn. del Lloyd Austriaco,del 1848, n. 173. 17. "Delle cognizioni di chimicaindispensabili alla scienza agraria." (V. Gazzetta di Zara, 1848, n. 11, 12.) 18. " AMetodoefficacissimo per la guarizionedi ferited'arma da fuocoagli arti complicatia fratturedelle ossa per cui pu6 risparmiarsenel'amputazione." (Art. pub. dall' Ecc. Governodel LitoraleAust. in Giugno,1848, e diramatoal personalcsanitario dell' armata Austriaca.) 19. " Antichelapidi salonitaneinedite illustrate." (Spalato, 1848, e 2da ediz. Zara, 1850, un vol.in 8vo.) 20. " Sulla importanzadella storianaturale e della economicorurale: Discorso inauLgurale,Zara, 1849, un op. in 8vo. 21. " De)la Topografladcell' antica Salona, con una carta topografica." (V. Alnali dell' Istitutodi corresp.Archeol., Roma, 1849.) 22. " Rapportogenerale degli Scavi di Salona,dalla loroprimo istituzione sino al giorlnod'oggi." (V. Annali dell' Instit.di corrisp.Archeol., Roma, 1850.)

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 266 Capt. R. F. BURTON.-On {he numismatist,and antiquary,who has travelledin England, and who has writtenhis travels. He kindlygave me a copy of his usefulstudy " Dell' antico Palazzo di Diocleziano in Spalato," &c., &c. (TriesteTip. del Lloyd Austriaco,1855), in whichhe has made sundry correctionsof Adams' the architect'sclassical folio,"The Palace of Diocletian," and of his "Antiche lapidi Salonitane inedite illustrate." In this volumethe inscriptions are translated,not merelycopied, after the lazy fashionof many authors,and the learned writer has freelyacknowledged the assistance of the celebrated Abate and ProfessorFurlanetto, and of his distinguishedfather, the late Carlo Lanza, a surgeon in the French armyof occupation. I can only expressa hope that his manuscript," Discorsi critici sulle anticheStorie degl' Illirici, dei Dalmati e dei Liburni," may soon see the light; and that the learnedauthor will republish,for the benefitof travel- lers,his valuable essay,entitled "Monumenti Salonitani inediti," printedin its Transactionsby the I. R. Academyof Sciences, Vienna, and in a separateform, also at Vienna, 185G6. We find the only notice of Illyrian Herakleia in the Peri- plus attributedto Scylax Caryandensis,and written-such is the differenceof commentators-betweenthe middle of the fourthcentury B.C. and the third and fourthcenturies A.D. (Muller). P. Farlati has charged the Greek authorwith inac- 23. "1Poche parole ancora sul colera,e specialmentedegl' insettiche furoni osservatigenerarsi nei cadaveridei colerosi."(D. Gazzetta di Zara, 1849, n. 124.) 24. " Sulle opinioniiriguardo alla contagiositadel colera: LJetteraal Sig. Red- attoredell' OsservatoreDalmato." (V. Osservat.Dalmato, 1849,n. 131.) 25. " Sulla Topografiae scavi di Salona dell' Ab. F. CarraraConfuLtazione." Trieste,1850, Ln opusc. in 8vo. 26. " L'Agronomoraccoglitore; Giornale ebdomadariodi economicorurale." Zara, 1850. 27. "KElementidi Mineralogiabasati sui nuovi principidi cristallografiae di chimiea,ad accerupagnatida practicheapplicazioni economiche industriali con noltefigure intercalate nel testo." Trieste,1852, un vol. in 8vo. 28. " Di alcune ricerchegeognostiche sulle formiazionidei dintornidi Zara." (V. Programmedell' I. R. GinnasioSuperiore di Stato in Zara, 1851-1852.) 29. " Nuove ricerchesulla formazionegeognostica. della Dalmazia: Rapporto rassegntoall' I. R. Direzione dell' InstitutoGeologico dell' imreroin Vienna." (V. Il Collettoredell' Adige Verona,1853, n. 4.) 30. "Elementi di Zoologia ad uso degli,I. R. Ginniasije delle scuole Reali Austro-Italiane,"2nd ediz. Vienna appresso Gerold,1855, un vol. in 8vo, con moltefigure intercalate nel testo. 31. " Dell' anticopalazzo di Dioclezianoin Spalato." Illustrazioinecon 12 Tav. orig.Trieste, 1855, uni vol. in 4to. 32. " MonuinentiSalonitani, inediti illustrati," con 10 tavoleoriginali. Vienna (per cura ed a spese,dell' I. R. Accademiadelli Scienze.) 33. " Sur les formationsgeognostiques de la Dalmatie: memoriapubblicata nel Bulletinde la Societe Geologique de France,nel f. di Dicembre,1855, con una tavola litografata." (For publication.) 34. "Discorsi critici sulle antiche storie degli Illirici, dei Dalmati c doi Liburni,"un vol. in 8vo.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Long Wall of Salona. 267 curacyconcerning the Narenta,but he is fullyrehabilitated, on thispoint at least, by the Abbe Fortis (ii. 152, Fr. ii. 208). As the passage of Scylax, though evidentlycor- rupted,and in places,with its "lacuneeet interpolationes," almostunintelligible, is of the highestimportance, it will be advisableto quote it at fulllength (cap. xxii. p. 28, "Geog. Greacimin." . Muller, Paris, 1860). IAATPIOI. MeTa Se At/3vpvoiVEc-Ov 'IXXvptoltMVoq, Kcal 7rapotKOV0tv ot 'IXXvptos rapa Od6aTrav jEXpt Xaovtag Trsj ara K'p Kvpavrr v'AX1ctv6ov v7ov. Kat 7roXt'EOT'TV 'EXXrvN9 E'Tavc, y ovo1.ta'HpacXeLa, Ka'tXqtwv. EMc 8eIcal o1 QXwro0a'yot KaXouJEvotBcmpi3apot ot'8ei ' eparar4wat BovXtvo' ('TXXto' ?) BovXtW&vo0WTEpLoves axxO. OV'Tot 8& q5aortv'TXov -rdv 'HpaKLeovJ av7ov KcaTOLKICrabCTI 86 Bap/3apot. KaTroKODO- & XeppovioTovAXiyq AcXacoy) Tp HeXo7rovvn7rov. 'A7rQ 8 Xeppov,rqov7rapacrr6vtov 4p06v [var. let. 'A7r0`S Xeppovr4oov (XZI V^io9) 7apa (Tretvet)@D rawvov 3pOov ?J rv 7rapo- ,oi3ouBovXvot. BovXtvo'83etffh) govos, 'I'XXvpLoV. Hapa67Xov 8'-rt Trfj BovXtvc X ypa'a 7epa paIcpaCt e7rb Ne'rTrov 7roTrap.v. Miulleroffers the followingLatinisation {" 22. ILLYRII. Post Liburnos sequiturIllyriorum gens, habi- tant que Illyrii secundummare usque ad Chaonianm,que est ex adverso Corcyra, Alconoi insule. Est que ibi urbs Greca, cui nomenIeraclea, cum portu. Sunt ibi etiamlotophagi qui vocan- tur Barbari hi: Hierostamne, Bulini (ilyllini ?); Bulinorum vicini Hylli. Hi Hyllum Herculis filiumsedes ipsis assignasse ferunt; sunt autembarbari, incolentes peninsulam paulo mino- rem Peloponneso. Post peninsulam vero (insula?) ornepree- tenditur quasi toenia recta; juxta quam accolunt B]ulini [Hoistenius"Post Chersonesum litus directumaccolunt Bulini"]. Bulini autem sunt gens Illyriorum. Prteternavigatioregionis Bulinorumusque ad flumenNestum est longi diei unius." Of Corcyra (KepKvpa peiXatva) we have no doubt. As regards Herakleia, the theoryof MIller is that the city's name found its way into the text as a mere gloss to the words'TXXoV Trv cHpaXeovq aivrovszaOCKz'a. It has long been suggested,and withmuch probability,that the 'IepoaT'u4vatis a corruptionof Ict8ePaTrrvat, the people dwelling upon the river of Salona, so well knownby Lucan's oft-quotedcouplet (iv. 404):

Qua maris Adriacilongas ferit unda Salonas Et tepidumin mollesZephyros excurrit lader. The name of the streamletis also writtenHyader, Jader,and Ider, the latterin the fifthcentury by Vibius Sequester (glossary): Pliny and theAnonymus Ravenno3 prefer Salon, and the Acts of S. DomniusSalonus; whilstCarrara (p. 1) believes it to VOL. V. U

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 268 Capt. R. F. BURTON.-ON the have derivedits name fromthe city. Similarly,according to Lovrich (" Osservazioni,"p. 11) the Rumin,Buda, and Grab streams,took theirnames from the villages throughwhich they flowed. This writerwould make Nastos, or Nastus, an Illyrian word,Na-sto (above a hundred), correspondingwith Cettina, alias Zentina or Zentena,because it was the chiefof a kundred castles or cities. Of the words BovXtvoiand 'TnXtvot,Muller remarks,appa- rentlywithout sufficient reason, "ejice vocem natam ex ditto- graphia nominisprecedentis." The Bulini are the BovXl,utE of Dionysius Periegetes (:387), who makes them conterminiof the Hyllic Region, and Eustathius explains that these " Bou- limeis" are also called BovXtvetsor Bovtvov'9. In the anony- mous poet vulgarlycalled Scymnus Chius (B.C. 92, if he be so old ?), we find,404- TovrTot9(Pelagones et Liburni) ovvd-7r-oVV'Yor? BovXw6vO

The learned historian Lucius (Giovanni Lucio), of Trai, would place the Bulini about the presentvillage of Bossiglina, commonlypronounced Bussiglina, an old fiefof the Bishops of Tratu,famous for fleeces. It is mentionedby Fortis as a pauper settlement,where the people cookedand ate arum,asphodel, and juniper berries.* We now approach the most debated part of the passage. Scymnus Chius (405) mentions the MelyaXfl Xepp'Vqo0-o 'TXXIK', believed equal (in extent) to the Peloponnesus,and containingfifteen cities inhabited by the Hyllaei. These people are describedas KeXTw6udE'Ovo9 (Etym. M. pp. 776, 39) and as Pelasgi by Niebuhr (R. G. i. 53). Pliny also (iii. 26) gives the Peninsula " Hyllis " a circumferenceof c. N. paces. Lucius of Traiu,would confinethe great HIyllicPeninsula to the lozenge-shaped,rocky-tongue of land projecting from Mossor,and bounded north by the Iader, and south by the Xarnovizza (the stream of the Xarn or mill), the latter unknownto the classics. This spear-headdivides the littlebay of Spalato fromthe long gulf of Salona, and its bold apex is the PromontoriumDiomedis, perhaps the later Fanum Dianae, a westernprojection of Monte Mariano,not "Marglian," so given by Wilkinson (i. 113). As Fortis very justly remarks(ii. 4, Fr. ii. 5 and 6) the Greekauthor must not be creditedwith the blunderof comparingwith the Morea a slip of land, a triangle * The shrubgrows wild all overthe limnestone formations of Istriaand Dalmatia. In the remoterparts, a wineis made fromthe fermentedberries, and the fresh fruitis used in medicine. A favouriteprescription against rickets in childrenis to pound to a paste in a stonemortar the freshly-gatheredberries, to mixwith aln equal part of freshbutter, and to applythis pommadeto the articulationsand the partsaffected, every night and morningbefore the child rises.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Long Wall of Salona. 269

only 12 mileslong by 5 of maximum depth,between the Islet of Rogosnizza and the Bossiglina village. The learned Abbe would,therefore, identify the Hyllic Chersonesuswith the rich riverinepeninsula fromthe mouth of the Titius of Pliny (iii. 24), the Titus of Ptolemy (Tt&ov 7ora,ov' Ic/3o0Xat,ii. 17, ? 3), nowthe Kerka of Sibenico,and the Tilurus,the modernCettina or River of Almissa. This tractcontains the " belle campagne" of Kniv,of Petrovopolje,and of Cettinaor Sign, and the site of Promona,the chief Illyrian settlementin the days of Augus- tus. This tract,he justly says, is still able to supporta score of cities. At the same time,he objects to the theorywhich wouldfind " Hyllis " in the rockytongue of Sabbioncello,which is distiinctlyalluded to by Scylax (cap. 23). Finally,we are tolerablysure about the Nestus riverand its accolo, the Nesti. The stream cannotbe the Titiusor Eerka, becausethe course of the Periplus is evidentlysouthwards to the Naron,which it names; the latter is clearlythe Narenta, or riverof Mostar (old bridge) with its " island 120 stadia in circumference,"still representedby the site of Fort Opus. Nothing,therefore, remains for it but to be the Tilurus or Cettina. Fortis holds the Nestians to be the people of the modern Primorje and the riverinesof Cettina. This coast, frontingBrazza,Lesina, and Sabbioncello,wascalled,in the middle ages, Parathalassia,which the Slavs translatedby the synony- inous Primorje(along-sea). In the days of the Avar invasioin it took the name of Pagania, from the Poganin or pagan Illyrico-Serbswho tenanted it; and, subsequently,the Arch- deacon of Spalato, GiovanniTomaso (in Lucio di Traui),*pre- servedthe barbarous" Maronia," also a synonymof Primorje. Prof. Lanza would place the Nestoi in the highlands of Pogliza or Poglizza; the ghats north of Dalmatia Proper, extendingfrom Clissa Fort (Spalato) to Duare Town,or between the debouchuresof the Xarnovizza and the Cettina streams. The name of this rugged oak-clad country,which has not, and, probably,never had a city,is by no means unknownto modern history; and its annals are so curiousthat I am temptedto a digression. According to Fortis (ii. 92, Fr. ii. 124-128), who ably sketchesthe pictureof the last century,this little aristo- craticrepublic, never containingmore than 15,000 souls, freed itselffrom the Porte,and, like its sister,Makarska, in A.D. 1646, threw itself into the arms of the "Serenissimo Governo," (Venice).

* Thomas (lat. A.D. 1200, ob. 1268) wrotethe " Historia SalonitanorumPon- tificumatque Spalatensium." JoannesLuciuLs has leftus the valuable "H istory de Regn. Dalm. et Croatie." He (lied at Rome in 1679, and his valuable MSS. are supposedto have been neglected.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 270 Capt. R. F. BURTON.-On the Societywas dividedinto threeclasses, which suggestthe Sixtiesand Four-candles,the Forties and the Twenties of aristo- cratic Guernsey. The firstconsisted of 20 nobleHungarian families,whom troubles had compelledto emigrate;the second were Bosniac (Christian)* nobles; and the peasantryrepre- sentedthe third. On theFete ofSt. Georget(Greek, April 19; Latin,April 23), an annual Zbor,Diet, or Assemblymet on theplain of Gattafor the election or re-electionof magistrates, each companycamping apart. The Veliki-knes(Knjaz, or Knight),the "Great Count" of Fortis,was alwaysa Hun- garian; theelectors or littlecounts, Bosniac nobles, represented the village communitieswhich theygoverned. Whilst the Governorwas being chosen,the plebs held theircomitia to electthe local chiefsfor the nextyear, or to confirmthose who deservedthe honour. The " FirstOrder " chosea captainand two procuratorsto supervisethe voting,and electionriots were common;whilst the "voto segreto" or "scrutin"was proceeding,some zealous partisan would seize the box containing theprovincial privileges (cassetta de 'Privilegidel Paese), whichthe law committedto the keepingof the Great Count, and runaway with it to thehouse of his favouritecandidate, in whichcase the latterbecame " belloed eletto." The difficulty ofthis proceeding was thatall theelectors might shoot at him, chasehim with their khanjars (long Turkish daggers), or throw stones,in whichexercise they are proficient,like the Syrians. The lawsof the Poglizanipreserved the rudenessof theages fromwhich they date. In cases of land disputes,the judge repairedto thespot, sat upona cloakor rugto hearthe plead- ings,and pronounceda decision, from which, usually, there was nioappeal. Aftera murder,the local court or governorand his notableswent to thecriminal's abode, and ate and drankhim out of "house and home"-a formof "dragooning"well- knownto all Easterns-and,finally, the Great Countand his comitypillaged all that remained.Formerly assassins were stoned,and thispatriarchal custom long endured in themodifi- cationwhich bears the name of JudgeLynch. For simple * Many Dalmatianfamilies derive themselves, truly or falsely,from the nobles of Bosnia, as we do fromthe Normans. Lovrich (p. 213) showsthe difficultyof genealogyby the systemof takingthe father'sname as surname; e.g. Philip,the son of Mark,would be Marcovich. He shows us the gradual growthof family names. "Quanto megio la intendonoi Dalmatini di oggi giorno (parlo di quelli,che non si vergognanodel cognomeSlavo, e che non lo Italianizzano)a non mutar cognomi da padri in figli,ma quello che lasciano i padri, traman. dano ai figli,ed ai nipoti." t An Illyrian proverb,cited by Lovrich (p. 78) is "Jurvevdanski Hajduki sastanski:" "George's Day; bandits array," because at that time the woods became leafyenough for ambuscades. If it rain on St. George's Day, cereals will be abundant.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Long Wall of Salona. 271 manslaughter,unaccompanied by atrocity,the "platiti ker- varmu" (lit. bloodspilt),* or blood-money,the Diyat of Moslem law was fixedat "' quaranta tolleri,"$40 or 8 zecehini (each = flor.475-5). The object of the fine was to prevent the criminalappealing fromthe decisionof the Great Count to the VenetianProvveditore Generale of Dalmatia. Those remnantsof the blood-and-ironages, ordeal by fireand boilingwater, were common,and bore the usual results,injuring and even permanentlymaiming the innocent,and allowing the sturdyruffian to escape punishment. One formwas worthyof Persia in the last generation: splintswere thrust under the finger-nails; the material was always the "ssapino" (Pinus maritimus),tbecause specifiedin the statutes,and the people wouldtolerate no innovation. The Poglizzan Morlaks were a robust and well-maderace, and Fortis gives thema good character,despite the patriarchal barbarityof their code. Sober and hardworking,they could boast- Durum a stirpegenus, natos ad fluminaprimum, Deferimus,suvoque gelu duramuset undis. They made a practice of bathing the babes in icy streams. These Morlaks were, and are, excellent irregulartroops, and theywere humane,hospitable, and friendlyto strangers,except when theirready suspicionwas aroused. They even refusedto speakof old documents,or to showinscriptions, lest the stranger whocould read themshould find treasure. Like certainidentical institutionsamongst differentnations, this warinessbelongs to a particular stage of development,and mustnot be attributed only to race. The village of Pirun Dubrawa (forestof Pirun) preservesthe remembranceof the god adored by the Slays of the city and provinceof Novgorod,before its conquestby Ivan Vassilovich, Grand Duke of Muscovy. As the old Pagans worshippedVid, so the Christian Poglizzans have an especial devotion for St. Vitus, and celebrate his festival by burning odoriferous woods roundtheir huts. Believing that if the perennialice be * Fortis,Lovrich, and Wilkinsolnall writekarvarina; thoughthe root is kerv, blood. Hence south of Cattaro the Kervosje,popularly written Crivosje tribe. The Morlaksfixed blood-money as high as 50, and even 60 sequins. t The tree flourishesall over the coast and the islands of Dalmatia,except wherethe windsare too strong: I need hardlysay that in the present state of civilisationno use is made of it but fuel. Yet it might take rankwith the growthsof the Thuringianmountains, which now supplythe " forest-woolpro- ductsand preparations,"cloth, yarns,waddings, oil, spirit,balsam, and soap, medicinal articlesso muchused in cases of catarrh,rheumatism, arthritis, and even paralysis. Dalmatia still importsthese articlesfrom Trieste, with the pine growingall aroundher house. Yet the Morlaks (Lovrich,p. 11) used " pece di sapino" in obstructionand phthisis.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 272 Capt. R. F. BURTON.-Onl the removed from their mountains, the Bora or north-easter would increaseto the ruin of their farms,they object to collecting and shipping it. Like all Morlaks,they are exceedingly jealous and yet they despise the sex, and hold the name so impure that it is never mentionedwithout a "saving your honour's presence." De prostite,moya zenal is the equivalent of the Maltese "Con rispetto (or con perdono) parlando, la mia moglie," as if his wifewere somethingimpure or offensive. Fortis seems to think this contemptjustified by the personal neglect of the women after marriage,but does he not confuse cause and effect? The churchesof Poglizza affectedthe Slavonian liturgy,and were served by the Glagolitic fathersof Almissa, who also labouredamongst the islandry. The militaryspirit is notextinct in a hardlyaccessible land, whereevery man is a man-at-arms. The mountaineers made a determined stand against the Napoleonic occupation in 1806. More than once theyhave threatenedAlmissa, and dischargesof cannonare the onlythings whichthey respect. Poglizza meridionale,the maritimelow- lands, are well-plantedwith fruit trees, and are nowwell known, becausethey supply Zara andSpalato with the best Marasca cherry, the basis of " Maraschinodi Zara." Returning to that debated ground, the Hyllic Peninsula,I may observe that some local antiquarieshave been so enthu- siastic as to findthe sepulchreof Hyllus in the finesarcophagus whichhas been placed for protectionin a chapel dedicated to St. Cajo the martyr. Its tripledivision, representing three of the labours of Hercules, is describedby every traveller. The learned Wilkinson,however, has neglected (i. 162) to notice, in the thirdor easterncompartment of the triglyph,the con- fusion of the Birds of Stymphalus with the Apples of the Hesperides. Prof. Lanza considersthe existence of Herakleia established by two coins in his fine collection; one with a metric diameterof 0.024, bears the head of a youth guardant right, and covered with a lion's hide (Herakles Imberbis?); on the reverse are the bow and the club, contained in a circle, and based by theexergue HPA. The second (0.013-0.016) bearsthe same obverse,but on the other side the bow and club are not in a circle,and the legend is HPAKL. Both have the letters * Lovrich (p. 164), in the days before Slav orthographywas fixed (1777), writes," S'prosetegniemnasce xene, nasce chieri, nasce rodizee" (con perdono, nostre moglie, nostre figlie,nostre parenti," &c.), whichmust be pronouncedItalianis- tically,and he tellsus that thereis no excusatoryformula when naming a man,the latterbeing nobler than " que 'sporchi,vili e sozzi animali,"as the Morlaks hold womento be. The idea is probablythe resultof an ultra-Spartanaffectation of manlinessand contemptfor effeminacy.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Long Wall of Salona. 273 well raised, and are tolerable specimensof the Greek type. According to Strabo (vii. I, i. 484), who notices the redis- tributionof lands every eighth year, the Dalmatfcdid "not use money,which is a peculiarityalso when comparedwith the habitsof theinhabitants of thiscoast; but thisis commonamong manyother tribes of barbarians." The Romans had no mintin Dalmatia, and only during the decline of the Byzantine Empire,Cattaro, Ragusa, and Spalato establishedtheir respec- tive "Zecche." Two medals of the Herakleian type are found in the I. R. Ginnasio Superiore of Zara, and are described by the Abate Simeone Gliubich (Numografiadalmata, in Italian, printedin vol. ii. "Archiv furKunde CEsterreichischerGeschichtsquellen," and in the Slavic " Arkiv za povestnicu Jugo-slavensku, Knjiga druga," Razdel i.) Heckel (Pt. i. "Catalogus Musei CaesareiVindoboniensis, numorum veterum") figures,in fig. 1, tab. ii., a roughlymade coin,with bow and club,and the exergue HPAK; and in p. 47 he ascribes it to "lHeraclea Taurica." He is supportedby Sestini (Monetavetus urbium, populorum et regum). On the other hand, the late Pietro Nisiteo,of Citta- vecchia di Lesina, a distinguishedstudent of Dalmatian antiquities, "proved," says Dr. Francesco Danilo (p. 173, "Pro- gramma dell' I. R. Ginnasio Completo di Prima Classe in Zara," 1849-1860; " Zara Tip. Governiale,"1860), "that these and other congeners belonged to the Illyrian Heraclea, men- tionedonly by Scylax Caryandensis,and placed on the sea-shore near the Liburni; in this opinionhe was followedby Gliubich." The only " Congenero" I can find noticed is a femininehead, coiff6ewith a bushel,and guardant right; on the reverseis a fish,naiant dexterwise; the diameteris 0.02, and the material copper,silver being the only othermetal used. Prof. Lanza's two medals were found near Spalato; but this provesnothing; coins travel as far as beads, Holloway's pills, and cowries. My old friend,W. S. W. Vaux, writesto me that eitheror bothmay belongto Heraklea of Thessaly,or Heraklea of Bithynia (Taurica), which had the same type and legend; but that,without seeing the coins,it is impossibleto assign the place of fabric-it can only be said that the Bithynianare the morecommon. To this objection Prof. Lanza rejoins,that the same type mightalso have been assumed by a thirdHerakleia, " Mentre sappiamocome gl'antichi popoli nella fondazionedi nuove colonie accostumasserotalvolta imporre a questoil nome ed adottare gli usi della madre patria." "At any rate," he concludes,"the find proves a commercialintercourse with the Greciancities furthereast." I visitedSalona forthe firsttime in companywith Professors

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 274 Capt. R. F. BUwRON.-On the,

Lanza and Michele Glavinic,Curator of the I. R. Museum at Spalato; the unaffectedkindness of this savazt, and his ready sacrificeof valuable time,have endearedhim to a numberof our countrymen. The act of escort was all the kinder in such weather; farfrom tepid was theturbid Iader, and no softzephyrs were the storm-winds. The shape of the old city has been comparedwith a rudder, the base eastward,and the narrow part to the west. Fortis (ii. 45, Fr. ii. 59) has noticed the corruptionof Caesar's text (Bell. Civ. ii. caps 8, 9) " Salona, in edito colle," when the " oppidum munitum" lies in the riverinevalley of the Iader, or Giadro; possibly,however, the conquerormay have includedthe fortof Clissa, the Andetrium of Pliny (iii. 26), which Fortis (ii. 48) writes "Mandetrium," and the Anderium('Av8ilptov) of Dion Cassius (lib. lv.). Here I saw at once the formof the double city, which, after the fashion of Athenae and Theboe,* convertedSalo, Salon, and Salona, into Salone. The base of the word may be Keltic, with the terminalon, or ona, signifyinga town-as Albona (high town) and Lisbona, to mention no others. The chronicler, Thomas Archidiaconus (nat. A.D. 1200), derives Salona from Salo, the sea; Rosacci from Calone (Joktan?) son of Salah (d-a'), son of Arphaxad (Genesis x. 25). Ortelli has doubts about the identityof Salona with the Salangon of Apollonius Rhodius (Carrara, p. 1). The Greeks, Strabo (vii. 5, ? 5, loc. cit.); Dion Cassius (lv. p. 586); Ptolemy (ii. 17, 4, viii. 7, 7); "Peeonio ix." (? Pteanius, Epitome of Eutropius); Procopius (di Bello Goth. i. 7); Zonaras (Chronicon. De Diocletiano), and othersuse 2%Xo, dctXw3v,:caXcw2mv, and acXc'Zvat/coXovi'a. Amongstthe Latins we findthe old marbles (Gruter," Inscript Ant. Amstelodama3,"1707); a leaden tube of the aqueduct; Pliny (iii. 22); Mela (ii. 3); Ilirtius (de Bell. Alex. cap. 43); the Anonymus Ravennee ("Europa," lib. iv. No. xvi.); Jornandes(De Regnor Success., cap. 58); and theold martyrologies preferring Salona, noe. The inscription referringto the road between Salona and Andetrium("Lucio InscriptDalm.," p. 34; "de Regn. Dalmat. et Croatihe,"p. 34; Caesar(" de Bello Civ.," iii. 8); Lucan (iv. 4); Vibius Sequester (Glossary); M. Aurelius Antoninus (Itinerarium) affect the less usual Salonae,narum. So statesCarrara (loc. cit. p. 1), but the use of the plural has evidentlyits reason. Colonia Martia Julia Salona, occursin an inscription(Gruter); Col. Jul. Salona on a medal of Claudius (Goltz de re numnmariaAntiq.); Col. Claudia Augusta Pia Veter.Salona, on a coin of Tiberius; fragmentsof potteryhave Salonas, and others,according to Farlati,

* AOtjlviiand Of'Mi(the Greek and theEgyptian), I need hardlysay, are both useed;but they would signify the old, or officialtown.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Long Wazllof Salona. 275

(Ill. Sac. i. 27) Silena. The names of the citizens and the adjectives are 2aXw3ves,preserved in the Acts of S. Domnius (Salones); Salonius (e.g. Salonia quercus) in Claudian and Servius; Saloneus in Priscus the Byzantine (A.D. 445); and Salonitesin StephanusByzantinus; themost general are Salon- itanusand Salonensis. Three other Salone are mentionedby the classics,one in Bithynia,a second in Gallia Narbonensis, betweenMarseille and Avignon, and a thirdinthe Gulf of Corinth. I am compelledto differwith my learnedfriend, Prof. Lanza, who,against the opinionof Carrara,believes the eastern to be the olderportion of the double city. The westernpart contains the little theatre,the Therma, whichhave evidentlybeen con- verted into an early Christianbaptistery, and the amphitheatre, which would hardly accommodatea large and opulent com- munity; its long oval is only 86 metres,and Mr. Paton reduces it to 126 feet. Again, the doublePorta Coesarea,the city gate" separatingold townfrom new town,has the towers (G) projecting eastward,and the re-enteringform is by no means the rule in Roman fortification.Carrara makes this royalapproach connectthe old Greek townwith Julia MIartiaSalona, as the western gate of Diocletian's palace, popularly called Porta Ferrea, united it with the suburb. Moreover,in the most massive and the earliest part of the north-easternmostangle (about F), near the Porta Andertiaof Carrara,I saw a Roman inscriptionbuilt up in the wall, and severalothers are recorded by the learnedAbbe, suggestinga comparativelymodern origin. Finally, the westernhalf has for its northernboundary the easternpart of the " Murazzo," or " Muro Ciclopico,"and, like the long walls of the Piraus, the westernpart of the venerable fragmentmay have been leftas a defence,cormmanding the high road to the Syracusancity Tragurium (Traui). Thus, I would believe,with Wilkinson,the long wall of Salona to be Greek, not Roman,and with Carrarato be pre-Roman.

PART II.-THE RUINED CITIES OF PHARIA AND GELSA DI LESINA. My convictionthat the long wall of Salona is Greek and pre- Roman reliesalso upon the fact that similarconstructions exist in the neighboarhood. Fortis (ii. 56, Fr. ii. 76) mentionsthem at Stobrez (Stobrech),east of Spalato, the Epetium foundedby the Siculo-Issani,the Syracusancolony which held the island now called Lissa: " Veggonsi ancora lungo le rive del picciol

* This double form gave rise to the Greek rvrXat,and we stillsee the grooves for the Cataracta,which the modern Italians call Saracinesco. This man-trap was a portcullis,let down fromabove like the gate of a sluiceby chains,and im- prisoningthose who had forcedthe outerentrance (IDennis, ii. 150).

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 276 Capt. R. F. BURTON.-On thleRtlined( Cities of

Porto riconoscibilivestigj delle antichemura d'Epezio, ch' erano fabbricatebensl di solidi materiali,ma senza quella squisitezza di connessione,che si ammiranelle fabbricheRomane." There are also ruinsof a Cloaca; the parish church,distant one mile fromthe fortwhich defended the land side, is built upon the old walls; anid the foundationsof a tower may still be tracedl. Fortis (i. 33) carefully distinguishedthese Greek remains from the Roman ruins, those, for instance,at the modernPodgraje (i.e. Pod-grada,under the city),the Asseria or Assesia of Pliny (iii. 21, Bohn i. 258). Here the walls are describedas 8 to 11 feet thick,8 feet high, and "lavorate a bugna" (draughtedand bossed). The Spalatinesspoke of classical ruinsin the islandof Lesina, which was undoubtedlycolonised by the Hellenes; and my attention was drawn to one not mentionedby any foreign traveller,when readingthe useful" Manuale del Regno di Dal- mazia (perI' anno 1873, compilatoda Luigi Maschele,Consigliere Imperiale," anno iii. Zara Tip. Fratelli Battara, 1873). The followingpassage occurs in p. 103: " GELsA-JJELS_A.*-XM - mentiantichi.-D)ue vetusti interessanti fabbricati trovansi nelle vicinanzedi Gelsa,entrambi posti sopra eminenzea mezzogiorno della borgataed alla distanzadi meno d'un miglio da essa. Il piiu antico e posto a cavaliere d'un monte. Questo edifizio,o a dirsi meglioquesto avanzo di antico monumento,viene com- unemente denominato Gor (read Tor) in lingua Slava. La fabbrica presentaun' opera di lavoro ciclopico,e fra i tre generi di tali lavori quello che veniva costituitodi massi regolaridi formaquadrilunga giA soprapposti uno all' altro senza cemento. Anche l'intero(interno ?) fabbricatoS di formaregolare, quasi pienamentequadrilatero. Per rimontarealla sua origine,bisogna ascendereai tempidi costruzioniciclopiche, che, come si sa, sono anterioridi piiusecoli all' era volgare. Lo scopodi tale fabbrica non puo ben determinarsi;isolata, posta sopra un monte di accesso aspro ed arduo,di non troppa estesa,non sapprebbesi conciliarein essa un' idea di abitato,e nemmenodi fortificazioiie, perch6senza argomentidi difesa. Forse pi\uaccettabile sarebbe l'idea, che si trattassed'un tempio antichissimo. Ad ogni modo il monumentorimane interessantissimoalla curiosita ed alle ricerchearcheologiche e storiche. L' altro monumento,a non molta distanzadal primoa levante,eretto anch' esso a cavaliere d'un' eminenza,chiamasi grad in islavo, che corrisponde a cittd o cittadella. Per il genere di costruzionecon pietre comunied a calce,per la distribuzioneinterna delle mura,per 1' esistenza entro tale circuitodi avanzi d'una Chiesa Cristiana

* The firstform is Italian,the second is Slav,pronounced Yelsha.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions P1lariaandi Gelsa di Lesina. 277

conentro una tomba,e perl'applicazionedegli avvenimenti storici del luogo si deduce,che la fabbricastessa rimontia tempi ben conosciutie storici,ed abbia avuto per oggetto la costruzione d'una fortificazioneper ritiratae ricoverodegli abitantida in- cursioninemiche nel sottostantepaese di Gelsa, che si sarebbe inclinatia riferire-anzichealle incursioni turchescheche non possonoascendere in su del secolo xv,-alle incursionidei secoli anteriori,e probabilmentea quelli dei Saraceni sopra l'isola di Lesina." Here, then,was at least one object well worththe traveller's study. My goodcfriend, Sig. Francescode' Vitturi,A. H. Lloyd's agent at Spalato, managed the transportfor me. MM. Paolo and Luigi Palese, civil engineersconstructing the mole and newport which are to connectwith the Spalato-SivericRailway, the firstof its kind in Dalmatia, kindly lent me their little paddle-wheeler," lessagero," and absolutely refused to be reimbursed,even for the expenses of crewand coal. On Monday, December28th, we steamed out, despite the furioussirocco or south-easter,one of the twin tyrantsof these seas, which was blowinggreat guns in theoffing. Spalato,by forceof situation, whichdetermines the rank of the world'scities, almost monopo- lises the rich trade of Bosnia, comprisedin cereals,hides, wax, and orpiment; silk, wool, and cotton," lihAfs" (bed-coverlets), copperpots and metals, iron, and perhaps gold and silver to come. The staplesof local productionbeing oil and wine,great effortsare being made to improve them; and the CEnological Society,worked by my friend,M. Aristide Vigneau of Bor- deaux,is doing much good, not only in making money,but in teachingthe peasant on the islands as well as the main,a new lesson,to prefer qualityto quantity. The portis notonlythe single settlementin the old kingdomwhich, I have said,shows any sign of progress,she is also thesole one that boastsof a trulybeautiful approach. Yet whenEurus is abroad,making the sea highand the currentslike mill-races, the noble amphitheatre, with its "peaksof lapis lazuli risingin majesticsplendour to the sky," the " impo- nentebaja," as Dr. Lanza justlyentitles it, is compelledto veil its charmsfromtheadmirer'seye.Thickwhirlingmist-clouds cap the straightdorsal lines of Mons Caprarius,the Kozj ak or goat-mountain of thetradition-loving Slav; the snowypeaks of MonsAuri (Mossor),which supplied theprecious meta? to theRoman capital;

Ibis litoriasMacer, Salonas; Felix aurifermcolone terree.

LLovrich quotes Pliny (Nat. Hist., xxxiii.4), " ut nuper in Dalmatia, prin- cipatuNeronis singulis diebus etiam quinquagena libras fundens, cum jam inventum in summocespite," and prefersthe Mossorof Promiliato that of Clissa as the originof thisgold.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 278 Capt.R. F. BURTON.-On theRuinzed Cities of and the regular,white-grey cone of Biokovo, still retainingin translation(Albicans or Albeggiante) the Albius and Adrius of the classical occupants. The rack cowers beforethe blast low enoughto hide the gap of historicClissa, and MIonteMariano gleams ghastlywhite against the angry purple sky. The deep blue of theAdriatic changes complexion to an unnatural ghastly green,upon whichwhite horses course,rear, and fall; and the bold and beautifuloutlines of the islandsfronting the bayletare hiddenby a curtainof cold greyfog. A few words concerningthe Dalmatian archipelago,and especiallythis sectionof it. Fortis justly describesthe islands as the sad remnantsof a land which has been tornby torrents, mined by subterraneanstreams, shaken by earthquakes,and finallysubmerged by a new sea. In vol.i. p. 18, he asserts"il mare guadagna continuamentesopra Zara; " and he givesmany names of submergedcities between Istrian Sipar (the Roman Siparum) destroyedin the ninth century,and the Bocche di Cattaro. Such are the floodedremnants of old Scandona,at the Lake Morigne,north of Sibenico; Nona, northof Zara, wherethe new settlement,called Privlaca by the Morlaks,is the Brevilacqua of the Zaratines,the Latin " Brevia aquse," or shallows; the ruined wall at the entranceof Makarska port, and the remnantsof Narona,now underthe swampsof the Norin river,the northern affluentof the Narenta. Even at Venice, in the sixthcentury, it was foundnecessary to defendthe mouthsof the cisternsfrom seawaterby raisingthe masonry. The whole coast of Primorje (Dalmatia Proper) has sunk,as may be seen by the sluggishness of thedebouchures affecting the climate,which once was so much praised. At Lissa Island thereare pisolithiccliffs, and at Lesina and other itemsof the Archipelago there are large tufaceous beds of rivers,apparently proving an original connexionwith the coast. Finally,at thePelagosa rock,called the "last Austrian ironclad,"a long dot of land seamostof the whole archipelago, where a fanal is now being placed, two tusks of an extinct animal are said to have been foundin the calcareoustufa, whose age, denoted by splendidVenuses, especially the Pectuncutlus (pelosus?) which still lives in the Adriatic,cannot be of great date. When Sig. Topich, mayor of Lissa, and now H.M.'s vice-consul,under whose charge is the historiccemetery of our seamen,was removing stone for the lighthouseplatform, he came upon an artificialcistern or gallery,containing human bones and othermatters, concerning which he has sentme notes and plans forthe Institute. Whilst the Dalmatian Sea is believed,since the days of the Paduan Vitaliano Donati (nat. 1717; ob.,Bussora circ. 1760), Manfredi,and Zenclrini,to have risen,that is to say, that the

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions .Phariaand Gelsadi Lesina. 279 shorehas sunk and is still sinking,it is generallyheld that the maritimepart of Adriatic Italy has risen,even since classical times. This seemsproved by the site of EtruscanAdria, which named the great gulf once upon the coast, it is now twelve miles inland; and the same is the case with Padua, R;mini, Ravenna, and Spina towns. This slow but persistentup- heaval suggests the shoresof the Baltic; and in neithercase can the phenomenonbe explained by the constant erosionand consequent deepeningof the sea, which Mr. H. P. Malet pro- poses to substitutefor secular upheaval. When, therefore,a modernwriter states of Dalmatia, 'e poi ben conosciutoche Ia spondanostra dell' Adriaticogradatamente aquista sul mare," he evidentlyconfounds the easternwith the westerncoast. The drownedcontinent of the Dalmatian shorefirst consisted of riverineand maritime plains; these were raised by the earthquakeand the volcano to hills,and, lastly, they sank below the sea-level. This theory,if correct,would give an indefinite lapse of ages for the formationof the archipelago. The older geologists made the depression synchronouswith the bursting of the Atlantic into the Mediterraneanbasin, and thence through the Bosphorus, events usually placed at the beginningof the glacial, which followedthe firstquateruary, age. Geologists still hold that in the second period of the stone-age Scandinavia, Jutland, and the Danish Archipelago were connectedinto a single continent. And the processof island-making still continues. The bold and serrated peninsulaof Sabbioncello is supposed once to have been the left bank of the Narenta embouchure; it is now connected with terra firmaby a narrow isthmus,and the sinkingof a few feet will reduce it to sea-level, causing total insulation. Between the Quarnero Gulf and Spalato the northernsection of islands and islets, often in double chain, subtend the shore-line,whose trend,to speak roughly,is fromnorth-west to south-east,and, as a rule,they present two parallel ridges ofhigh ground,as if a continenthad been shatteredinto a hundredfrag- ments. The double chain of sunkenmountains, broken by bays, sounds,and inlets,seems, when viewed from an elevatedpoint of the coast,to fitinto the terrafirma as if once joined on to it. But aboutthe parallel of Punta Planca, thewesternmost projection of Dalmatia,the shore-linebends into an importantchord concave to the north, and here the chief islands,Brazza, Lesina, Lissa, Curzola,and Lagosta changefromadiagonal (north-west-south- east) rhumbto a parallel of latitude,the lengthof all fourbeing disposedalmost due east and west (magnetic). Finally, further * A few,but very few,have called the Adriatic fromAdris (Hat or Hatri, hod.Atri) of Picen'n, themodern Abruzzi.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 280 Capt. R. F. BURTON.-On theRuined Citiesof

south,as far as the Bocche di Cattaro,where the archipelago abruptlyends, the groupsre-assume the diagonal trendof the northernsection. We had some grief in traversingthe Canale della Brazza, betweenthat island and the continent,and again the little steamerwas heavily shaken by a cross sea in the Canale da Grecodi Lesina, which has Brazza to the east. Fortunately, however,my kind-heartedand obliging friend,Cav. Pietro G. di Leva, an old Pacific voyager,now port-captainof Spalato, had taken the precautionto send with us Gospod Dumantich, his chief and most experiencedpilot. After four hours, we foundourselves running S.E. in smoothwater, with Port Pelle- grino to starboard,and the highlands of Port Kubal to port. At the bottomof the bay lay Cittavecchia,or Starigrad,and hereI was hospitablywelcomed by Capt. Pietro Ivanisovich, the PodestAor mayor. The island of Lesina, says Capt. GiacomoMarieni (pp. 331- 349), in that fine folio the "Portolano del Mare Adriatico," (Milano,Dall' I. R. Stamperia,1830) is one of the largest and the most populous of the Dalmatian Archipelago. Its length from east to west is thirty-seven(Italian) miles, whilst the breadthvaries fromtwo to three. Fortis gives these figures forty-four,and a maximumof eight. Until the early part of the presentcentury it was well wooded; its forestsof the Pinus maritimus,which Linnacus ignored, combined with its peculiar, long, narrow shape, gave rise to its two known names, the originalLiburnian having whollylapsed into oblivion. The Greeks called it IIdpos or adposX, which we find in Scylax Caryandensis(chap. 23) aviOa ryacpecrrT veo Mapos, Vr?7o-o 'EXXev'W, Ica4 '"Iv'l 9oo, Kal7rcXev9'EXXevl&e9 aiTat. AppoloniusRhodius, describing the passage of his knight-errants, the Argonauts,terms it ItTv.etca,or pine-island(w7-ITv = pinus) in this verse (iv. 664):

"Iaa-a Te, Zv7/fceXacot ica\ 11weprT\I16Tv&eta. UJponwhich the Scholiastremarks, A t'7OVpV0o eOvo9 oiicovVreq \ ravras ra v vs,"Iaav . . . . /zeO'&a EL9 KE'p1vpav j'XOov. T73V8FEl\rvovYo-a-av 1KaXoV/JE'lNV ILTVetav e7Trev'O,uqp(P evopeq, Someauthorities have applied " Pityeia," withoutsufficient reason, to the woodedrock of Sant' Andrea, wherepitch was drawn by incision. "IDyskelados," accordingto Fortis (ii. 163; Fr. ii. 222), is the title of Issa (Lissa); the "harsh-sounding"`is generallyunderstood to be Cratia (Kparta), Crathis (Kpa'Os), Bractia(Bpec'Tta), or Brattia,the BaprTcOof C. Porphyrogenitus, and the modernBrazza, still so heavily visitedby the howling Bora (north-easter). Pliny (Nat. Hist. iii. cap. ult.) applies to

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions .Pharia aindGelsa di Lesina. 281

the fiftyislets and rocks of Sibenico the collectivename of Oeladussoe,supposed to be derivedfrom Dyske'ladoi; andcFortis (i. 170, Fr. i. 235) thus amends the corrupted text, "Nec paucioresTrucones (insulae) Liburnice. Celadusse contraSurium (ZuriIsland). Bubus (Bua ofSpalato), et caprislaudata Brattia." Scymnus Chius (Periegesis, 1. 427) adds another detail: Mapo9 8' rov{rcwv(the Illyrians) ovIc &7rcoOevxe'q4v Nro-osllapt'aV IMGT'fS 6W. Diodorus Siculus (xiii. 3, 4) relates the decree of the oracle which determinedthe Parian emigration,and dates the foundation of Neo-Paros fromthe year "when the Elhans were cele- bratingthe 99thOlympiad (B.C. 385). The new-comersexpelled the barbarians,who took refuge in a very strongvillage, and quickly foundingtheir city near the sea, walled it round and held it for 166 years. Strabo (vii. 5, s. 5) refersto it as 'HEapo9, Ilipo9BXeyopev'Y prp6Tepov.It is the iapia of C. Ptolemy (ii. 16, s. 14), who applies the termto isle and capital, and the dapa of Const. Porphyrogenitus(De Admin. Imp. cap. 36). The Romans,as we learn from Pliny (N.H. iii. 30), preferred Pharia, probablypronounced Pharia, and the Slays, who con- vert Ph and F into Hv ( = Kh in such Persian words as Khar and Khwar), have retained ivar, evidentlyfrom Phar(ia) or Far(ia). "The name is given in ancient documentsto the islandand, after the foundation of Lesina town,to Cittavecchia."t Therecan be no reasonabledoubt of this identification when we inspect the coins turnedup at Cittavecchia. In the days of Fortis (1772) a single specimen was known. Prof Boglid (loc. cit. p. 18) mentionsbut threesilver in 1873, one described by Sestini,a second owned by Sig. G. Macchiedo,and a third in his owncabinet. The "Programme" beforealluded to describes (p. 171) fivecopper medals in the Gymnasiumof Zara, not to speak of the manypreserved elsewhere. The characteristicsare the virile heads,bearded (Jove ?), or imberb,nude, laureatedor crowned,and guardant dexter or sinister. The reverse often showsthe goat (Caprone), derivedfrom Paros of the Cyclades, and No. 5 bears four rays above it. Prof. Lanza's copper specimen seems to have an olive branch over the animal's crupper. The exergueis P, and the diametervaries from 0.019 to 0.024. Prof. Ljubi6 (or Gliubich), of whommore presently, declares (p. 8) that the differenttypes found at Cittavecchia exceed one hundred,a numbersurpassed by few ancient cities. * Bostockand Riley (Bohn i. 266) thus pervertthe passage: " Opposite the Liburniare some isles called theCratece, and no smallernumnber styled Liburnice and Celadussu. Oppositeto Suriumis Bavo, and Brattia fanmousfor its goats." t This statementof Prof.Ljubio (p. 33) is contradictedby Prof. Bogli6 (p. 34).

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 282 Capt. R. F. BuRTON.-On theRuined Citiesof

Ile gives as the chiefobverses the heads of Ceres and Homer; the reversesbear the serpent,the goat, the pine, and the two- handled wine-jar, emblems of agricultural prosperity; the exerguesare mostly!lA and !AP, sometimeslAPL(2N, and HA (Bogli6,p. 18) in only one specimenfound at Lesina town. According to Sig. GiacomoBogli6 (p. 31, Studj Storicisull' Isola di Lesina di G. B. Prof.nell' I. R. Ginn. di Zara: Zara Tip. di Gio. Woditzka, 1873, Fasc. 1, pp. 5, 31), the Slavs originally called Pityeia "Lisna," and subsequentlyLesna, Liesna, Liesena, and Lesina. He quotes the Lesignano, Vinc. Priboivo (Venetiis, 1525), who terms it " Lisna," adding, "corruptiusautem Lesina dicitur." Yet Prof. Ljubi6 (p. 30, "Faria," &c.) notes that in a documentdated June 19, 1103, by Koloman (A.D. 1095), son of St. Ladislaus, of Hungary, and quoted by Lucio (vi. c. 4), and Farlati (iii. 164) has "et villam Lesina Stolez." Ambrogius Calepinus (p. 226, " Onomasticonpropriorum nominum. Basilie," 1598) speaks of the " insula quoe nunc vulgo Lisna Illyrico sermone dicitur." Prof. Bogli6 would derive Lisna (the woody) from Lies (a wood),quoting the Lexicon Serbico-Germanico-Latinum (p. 328, Edidit Vuk Stef-Karadschitsch,Vindobonoetip. Mechit., 1852) " Lies (u lIrcegovini) Silva. Blagi jezika Slovinskoga (in the blessed Slav tongue), (p. 232, Laureti, 1694) Li.s,Lies. To jes driva za poslovati (i.e. wood to work). . . . Lignorum apparatu6s.In Bosnia there is a Mount Lisina and a River Liesnica; Serbia owns three villages called Ljesnica, and at Rudine, near Cittavecchia di Lesina, we find the Valle LesnA and the Punta Lesnirat. As the Slavs make littledifference in the articulationof the i, the e, and the diphthongie, the name was writtenin the public documeentsof Venice indifferently Lisna, Lesna, Lesina, and Liesina. Finally, in the early 19th century,the shape of the island suggestedthe modern and popular Italian form "Lesina," meaning a "cobbler's awl," and the trueterm (Lisna) lapsed into oblivion. According to ProfessorBoglic, the island aboundedin pre- historicor proto-historicremains, whose originhe wisely hesi- tates to determine. His descriptionis as follows(pp. 7 & 8):- "iMonumenti sepolchrali ci restanosull' isola, e moltiintorno alla cittAdi Lesina, i quali senza dubbio rimontanoall' epoca seconda dell' eta di bronzo. Parecchi ne furonoaperti, sono dodici anni circa (about 1861), per cura dei Signori Gregorio Bucic e Pietro Bogli6. Erano essi formatida grosselastre di pietra non isquadrate,lunghi dai tre ai quattro piedi, larghi due, e a un dipresso altrettantoalti. Enorme la grandezzadel coperchio,e di alcune dele lastre laterali poste sempre per coltello (on edge).

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Pharia and Geksadi Lesina. 283

"Ecco quello che mi scrivevail Sig. GregorioBuci6 riguardo il coperchio-- una sola era la lastra ciheserviva di coperchio,e sempremassiccia assai e greggia-una volta massiccia tanto dhe convienefrangerla col farvicadere dei grossissimisassi, essendo stato impossibilesollevarla a braccia, sebbene molti operai 81 lavorasserointorno.' "E siacomenon poterano estrarsi sopra luogo, ch a cib non prestavasila natura del terreno,fanno suporresforzi e fatica somma,perch6 potessero trarsi lassALsenz' ajuto di macchine. Queste tombecostruite sopra terra,senza calce, sempre in cima di qualche collina, protetteerano dalle ingiurie del tempo e degli animali da un tumulo di sassi, alto finea dieci piedi, di una periferiache talvolta aggiungeva i cento,e la cuii somita finivaper lo pitt in pietre di considerevolemole. Le tombe non giacevano mai nel centro del tumulo, ma erano poste alquanto a levante. In una si trovo conservata una parte di utncranio, e dei carboni,in un' altra podhi carboni,due vasetti di terracotta; taluna conteneva ossa frammistedi adulti e di bambini. Si raccolse un fusodi bronzo,e dello stesso metallo uniafibbia a semicerchio,un anello, un amo, alcune di quelli spirali che sono caratteristichedi. quest 'epoca, e pezzettini di ambra. "I vasi, gli oggetti di bronzo coi pezzetini di ambra allora trovati,Si conservanoa Lesina dal Sig. GirolamoMachiedo, ma le ossa del cranio,ed iAfemore furono riseppelliti,e forsenon eranodi minorinteresse per la scienza. II Dr. Francesco Uinger, Professoreall' UniversitAdi Vienna, rapito troppo presto all' affettoand alla gratitudinedei Lesignani, avendo esaminato nel 1864, gli oggetti di sopra accennati,anch' egli i giudico dell' eta del bronzo. "Qualche sepolero era vuoto, ne pare che prima sia stato aperto,perche non si vide alcuna traccia di lavoro untorno al tumulo, ne appariva che le lastre delle tombe fossero state smiosse. In questo non mancava il terricio (black degraded earth),prodotto dall' azione dell' uiiditA e del tempo sui resti aniimali,che vi erano stati deposti. In nessuna si trovarano oggettidi ferro. Molti anni prima alcuni agricultori, spinti dalla speranza di arrichirecon tesori nascostivi,spianarono del tumuli, e diseppelivanotn grandevaso di terrasenza manico,ed un martellodi bronzo,scarso compensoal lungo e penoso lavoro. In uno dei sepolchri da essi aperto, tanti erano i carboni accumulati,da dover sospettareche entrosia stato abbrucciato il cadavero. La tombanon era semprelunga cosl da contenereil cadaverodisteso di un uomo,e siccomeda ossa trovatevisenza traccie di fuoco,si vede, che non tuttii cadaveri venivanoarsi, conviencredere, che in qualche caso si ripiegassero; tanto piil VOL.V. X

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 284 Capt. R. F. BuRToN.-Ow-the RBtined Cities of

che la periferiadi un cranio, desunta dalle ossa parietali e dali' occipitalibene conservati,e la lunghezza di' un femore mostrarono,dhe gli uominia cui appartenevanoerano di statura anzi alta, che mediocre,e di grosse proporzioni. La forma piramidaledei tumuli,il colore,e la grandezza delle'pietre che ne costituisconol'esterna corazza, il sitoelevato, sterile, sassoso, su cuiis'innalzavano, Ri fanno tosto riconoscere e distingueredalle macerie(heaped stones)-diegual mole costruitedagli agricoltori -collepietretratte dalle pendicidellecolline ihedissodano. Devesi deplorare che quando si spianarono i tumuli, nessuno si sia pensato di studiare pithattentamente questi monunenti dell' ot&preistorica. Se con questo scopo e con maggior cura fosserostati eseguiti i lavori, si sarebbero ottenutepii precise indicazioni,e forse tra i vari sepolcrisarebbesi potuto rilevare un grado diversodi antichitA,." The passage is interesting,because it shows that Lesina contained a styleof tumulararchitecture dating before the Iron age, and thus equal in antiquity to the oldest Etrascan remains whichlie on the confinesbetween the Bronzeand the Iron. Oittavecchiade Lesina, the " Civtas Vetus," which the Slavs, rend:eredby " Starigrad" (old fencedcity), opposed to Civitas Nova, Novigrad,or Lesina town,is the usual Veneto-Dalmatian port-town,a gatheringof big, dull houses ri'sg fromnarrow alleys,which were neitherpaved nor lightedtill thereignof the presentPodestA. The -redeemingpoint is the neat riva or quay of cut-stone-a luxury found throughoutthis seaboard from Sibenico to Cattaro,and still wantingon the Surreyside of the Thames. The little port requires a prolongationof the rudi- mental mole,at whose base stands the Saniti-box,as westerly winds drive,at times,the billows right home,to the imminent danger of the shipping. The fourchurches, including S. Pietro, the Dominican monastery,do not exceed the usual allowance, and the Parrochiale and ex-cathedr-al(?)* of S. Stefano, -whichhas a cachetof its own, bears over the entranceof the belfrythis barefacedinscription in thebaldest Latin- Dederunthujus primordiamolis de nenlibhusurbis reliquie, Et qutedederat gressumn in urbemjanua, nunuin templisacrarium. The people,who numbera maximumof 4,000 out of a total of 14,000 to 15,000 islanders,regret that their" city" was not built on the slopingground a little farthernorth, where the * For the origin of the bishopricat "

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Pharia and Geha di Lestina. 285 drainagewould have been better. High-spiritedand industrious, stoutfishermen and gallant sailors, theyleave -their women to till the grou.nd- Fweminapro lana cerealiamunera frangit Impositoquegravem Vertice poPtat aquam- and theytake a pride in calling themselvesthe English of Dal- matia, a distinctionwhich they amply deserve. Their island yields grain enoughfor five mont-hs only; the othereight must be supplied with bread fromthe Danube and the Black Sea.* The chief local cerealis barley,the rainsnot being heavyenough for " fromenton"or maize. Wine is abundant,alcoholic enough to suit the English market,and much resemblingthe Lavradio and the inferiorgrowths of Port. The-olivethrives everywhere, and at this season the streetsand stairs are -rankand slippery with oil. The Chambers at Vienna, influencedby a great monopolistcompany, have lately done an unwiseand unpopular thing in refusingan annual sulbsidyof 20,000Oflorinsto an inter- insularline ofsteamers. The difficultyof intercourse here causes not only inconvenienceand loss of time,it also affectsthe trade of manycommunities, and renderstheir progressand develop- mentnext to impossible. I at once inquiredabout the so-called " Muraglieciclopiche," a term adopted throughoutDalmatia fromDodwell (" Views and Descriptions of Cyclopean or Pelasgic Remains," &c., a paperposthumously affixed to his "Tour in Greece"); fromPetit- Radel, who proposed the theoryin 1829, and who developedit in 1841 (" Recherchessur les MonumentsCyclop4ens, et descrip- tiondes modelesen reliefcomposant la Gallerie P6lasgique de la BibliothequeMazarine," Paris); and by Cesare Canta in 1846 ("Dei Monumentidi Archeologia"). A local antiquary has determinedthe walls ofPharia to " belong,without doubt, to the second species of Pelasgic construction,"the firstbeing "irregular polygons,uncut, or cut afterthe Lesbian rule,concerning which the author, 'De Mirabilibus,' speaks." Sig. Girolamo Budrovich,formerly Sindaco, and now Agrimensor(land sur- veyor) of Cittavecchia,has succeecdedin outliningthe limits of the old city-of course, only the officialportion-an, oblong -measuringabout an Italian mile in circumference.The dimen- sions of the enceintewere 36 feetin thickness,and the height was conjecturedto be about the same. Possibly the normwhich Fortis applied to Aseria (Podgraje) has been fittedto Pharia. The fragmentsbest preservedare in a cellar (Cantina di Gram- * The Euxine is in everymouth; for instance,Fortuna nel G-olfo(Adriatic) burrascain Mar Nero; and duringa stormthey will exclaim," Poverettinel Mar Nero!"

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 286 Capt. R. F. BURTON.-On) theRuined Citiesof

motor),where, however, they are almostconcealed by gloom and lumber; outside the town,about 10 metresrise above ground near the Casa Domehich; and the foundationsremain in many places uninjured. Hence a descriptionand, I believe,a plan were publishedin the InstitutoArcheologico of Rome between 1840 and 1848. Accompaniedby the gentlemenof the town,I visited these interestingremains, and found the masonryto be pseudo-isodomic and isodomic(regular horizontallayers), parallelopipedons of considerablesize, cut fromthe limestoneof the neighbouring hills, hammer-dressed,and fitted upon one another without cement. I nowhereremarkled bossed and draughtedstones, but Sig. Budrovichassured me that theyexisted near the city,and he presentlyobliged me with the followingthree specimens. The measuresare in Viennesefeet and inches.

=-8"

FIG. 1. FIG. 2. FIG. 3. The centralprojectioii above the draughtis 4 inches in fig.1; 1 inch 6 lines in fig.2; and 2 inchesin fig.3. The good Agrimensorwas also kind enoughto informme by letter(Feb. 4, 1875), that about an Italian mile outsidethe town he had foundmasonry which appearedto be of the oldest date. Fortis (ii. 176) had previously copied a Latin inscription "'perhaps a mile fromCittavecchia." There has been since the early 14th century-probablylong beforeit- and therestill is, a mightyfeud betweenCittavecchia and her south-westernneighbour, the "HHaupstadt " of Lesina, about the rightof representingthe ancientPharia of Demetrius of Paros. Fortis (ii. 175), guided by a local tradition, would place the site some two miles furtherinland, where ruins arc also found; these evidentlybelong to some forgottenitem of the fiveclassical towns. In our day the leaders of the war are the two followingliterati. Professor Boglic (p. 26, loc. cit.) rather "'trinms"between the two. According to him,the Paros built by the Parians in the 4th centuryB.C., and utterlydestroyed by the ConsulLucius Emilius Paulus in B.C. 219 (Livy, Polybius, Dion Cassius, Florus, Appian, and Justin),occupied the siteof modernLesina. The neo-Pharos, the ve'09 lapos of Scylax (??); the Ager Colonicus of the Romans; the Oppidum Pharia of Pliny (iii. 26); and the Pharia city of Ptolemy (loc. cit.),' which * Consult the " Faria," &e., of Ljubi6 (pp. 25-28), forproofs of the Roman epoch betweenB.c. 219 and A.D. 639.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Prhatiaand Gelsa di Lesilna. 287 lasted till the end of the 5th or the early 6th century,when the Slavs foundit in ruins,was built whereCittavecehia now stands. On the other side, the doughty abb6, Dr. Simeone Ljubid (Gliubich), at present professorin the I. R. Gymasitm, and Curatorof the Museum of Agram (" Faria Citt&Vecchia e non Lesina," Zagabria. Carlo Albrecht,1873, an octavoof 68 pages, printed with the aid of the municipalityof his native town), soundsno uncertainnote. The followingis the merestoutline of his arguments(pp. 7, 8) in favourof Cittavecohia;of course most of the statementsare traversedby Prof. Bogli6. The latterlocalises near Lesina townthe attackof L. Emilius Paulus, who razed the city which the Parians built in B.C. 385. He mentions (p. 15) ani ancient mosaic two feet below the surfaceof the Bishop's gardenins modern Lesina; anothernear the Mandracchio,or dock port; a fragmentof granite column about two inches long; a pair of the sepulchranlamps called by the Italians " lume eterno"; and a multitudeof coins, especially twenty-twoof Ballmeus,found in late yearson the slopescrowned by the Forte Spagnuolo. a. The existenceof the old walls of Cittavecchiaresembling the Paleokastra of Albania, &C., &C. I. The many Greek inscriptionsbearing the initials or the words 46psos and 4kapkowv,and showing the old democratic republic to have been governed by Archontes,Prytaneis with theirllpvTave 'ov; a Senate (BovXr) and a Scribe (eypaCTC'ErEo9) in the public 2Erarium(ev T-' y.oo4). Another inscription, "di argomentoimverecondo, is a Psephisma,or decree regu- lating the public prostitutionwhich seems to haunt everyport. a. The multitudeof Pharian coins inscribed OA, !AP, and OAPI2N, andbearingthenameofBallous (BAAAA, BAAAAI, BAAAAIOT, and BAAAATOT (?)),at firstArchon, and lastly tyrant(BaotXev's) * of his native Pharos. d. The numberof other classical coins which have come to light. e. The incised gems and cameos. f. The sarcophagi "of primitiveor Greek style,almost in- variablyanepigraphic," containingmortuary objects. One of these,discovered in 1869, when prolongingthe riva southwards, yieldeda gold ring with incised stone,and an exquisitelymade chain of the same metal; both are now in the Museo del Tri- regno of Agram. " Some of the mortuaryurns much resemble in shape those of Issa, which are undoubtedlyin the Liburnian style." * Prof.Ljubi6 (p. 9) givresBaarrXsor, but as there is no accent,the omission maybe a misprint. He is also the authorityfor the form BAAAATOY (p. 9). Pwof.Bogli6 says that the coins bearingBA ZA are veryrare (p. 18).

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 288 Capt. R. F. BURTON.-Ofl theRuined Citiesof

y. The marblebas-reliefs and similarobjects, the threeprin- cipal being-1. A mortuarygenius, with reversedtorch in the right hand, and in the left a bunch of grapes; it is a coarse production,placed on a pedestalof cut stoneat the cornerof the little square frontingthe parochial church. 2. A Milo of Crotona, in mezzo relievo. 3. A marble slab, showing an archaic Grecian galley, steeredby a pilot. The last is high up in the Campanile; it is noticedby Fortis (ii. 175, Fr. ii. 238) as " un bassorilievosufficientemente ben conservatoin marmo Greco,che rappresentauna barca a vela, col timonealla destra della poppa, ed il piloto che lo governa." On the other hancdLesina, says the Abate (p. 11), shows no signs of those " primitive (Pelasgo-Illyrian ?), Greek and Roman monuments. The town (whose origin is lost in the gloomsof the middle ages) is stuckagainst a nude and savage rock; it has no agro (open ground) even to the south,and the fewyards of dry surfaceforming the actual square,and bearing, amongst other edifices,the Duomo and the Episcopal Palace, were coveredby the sea withinthe memoryof man; in ancient timesthey must have been deep water," &c., &c. Finally, the eruditeprofessor quotes a host of authorities,who reject Lesina and supportCittavecchia. He has, however, omittedto notice that the position of Cittavecchiais exactlythat of ancient Salona-a recess,a sinus openingwestward with higher groundto the north and south-, thus defendingit from the Bora, as well as fromthe Sirocco. Lesina, seated like Spalato, on the very edge of the mainland, faces the southwith a much greateramount of exposure. All the Cittavecchianideclared, to my unpleasant surprise, that they had never heard of " Cyclopeanstones " at Gelsa. I determined,however, to judge for myself,and, kindly guided by the Deputato di Sanitg, Sig. Gio. Actinovich,I set out to cross the island. The distance from Cittavecchia to Gelsa measures upon the map 4,000 Viennese klafters(= 8296.82 English yards, = 4.714 statute miles). Under the justum et per mundumapprobatum regimen of the SerenissimaRepublica, to which Lesina was yielded in A.D. 1420 by its last Count, Aliota Capenna,*this highway was a good calfada of cut stone and lime; now it resemblesnothing more than a torrentbed, except during heavy rains,when it becomesa shallowtorrent, a * Accordingto Prof. :Boglic (p. 50), Petrana, the Venetian Count of Zara, compelled Lesina to undergo the yoke about A.D. 1144. There was a second renditionin A.D. 1278 (p. 73); thefinal subjection after the last swayof the Hun- gariansceptre was in A.D. 1420 (p. 103). The threecommissioners who acted in the name of the Reggenza (count and judges) ; the nobilityand the university (popular assembly),were Giovanni Ozor (de Ozoris), Vital di Silvestro,and Vito di Tomaso (Jan. 3, 1421).

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions P4aria and GCelsadi Lesina. 28

fiumara,or, in local dialect, a "potok." It is everywhere enclosed,first by tall masonry,and afterwardsby dry walls. The almostlevel surfaceof the ground,a shallowprism, with highlandsto the rightand left,would representthe old Pharian ager. Its clothingof rich,ruddy clay, the produceof degraded chalk and ferruginouslime, is scatteredwith water-rolledcbits of " brecciacorallata," from the summitsof the southernrange. Fortis supportsthat this materialis the " Tragurianmarble," so highly prized by the Romans-angular fragmentsof white calcaire encrustedin a cement of petrifiedochraceous earth, susceptible of a high polish, and often confoundedwith the African. He never found amygdaloid, nor pudding-stone, but always large, flat, angular, and irregularpieces, inter- rupting the continuityof the texture. These breccias, being frequentlymet with in situ,upon the summitsof the Dalmatian ranges,confirm him in the belief that the islands were once part of the continent. I remarkedthat in some cases the white marmorinelimestones set in the blood-redpaste were rounded like true conglomerates,and I heard of a vinous colouredrook, probably the "Rosso da Cattaro" of Venice, which may be seen in the pavementsof streetgand churchesat the place which gives it a name. The routebecame moresandy as we approachedGelsa. The groundon both sides, broken into hill and dale, was enlivened by white-churchedsettlements, which are rather miniature citiesthan villages, and nowherein Dalmatia had I seen dimen- sionsso considerable. Thus we read in 1798 (" Engel, Geschich. v. Dalm. Allgem. Welthist." Halle, xlix. 228), " Verschiedene Dorfer derselben (Lesina) verdienendem Namen von grossen Flecken, und sind stiirkerbewohnt als viele kleine Stiidte." The phenomenonis doubtlessdue to the comparativewealth and abundanceof the island, which bears besides wine and oil, figsand almonds, saffronand honey. The rakia (raki of the nearerEast) is particularlygood and plentiful; the aloe flowers as in Greece,and at Lesina the fibreis made into fancyarticles, purses,and nettings. Although the woods have been unwisely thinned,and re-forestingis requisite,palms and carobs,oranges and mulberrieseverywhere thrive. Sheep-wool and cheeses have taken the place of the salternswhich, in 1772, formedthe chief local industry,and the habit of long voyages has not abolishedthe fisheryof anchovies,mackerel, and sardines,with which Lesina, in the days of Bosching, suppliedItaly, and even Greece.* * In 1861 a certain Sig. Carlo Warhenek began to cure sardinesin oil at Fiume, and presentlytransferred his establishmentto G0elsa,the centreof the Dhalmatianoardine-fishery. But the want of transportand the exsesive taxeuou

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 290 Capt.R. F. BURTON.-0n theRuined Cities of

On the right I remarked the settlementsof Dol, Sfirze (Svirze ? in Ital. " Sfirce"),* ancdVerbanj, whilst to the left rose the finelarge houseswith which the littleport of Verboska was adorned some threecenturies ago. Half-waywe passed a dilapidatedlittle chapel, which my good guide called S. Cosmo, and whichappears as S. Vito in the officialmap of the Austrian Empire (Special-Karte des K6nigreichesDalmatien, &c., &c., &?., von dem K. K. Militirisch-GeographischenInstitute in Wien, 1861-1863). Beyond Verbanj we fell into the valley of the " Potok," as the streamletof Gelsa is genericallytermed, and crossinga big stonebridge, we presentlyfound ourselves in the little town. Here we were met by the PodestA,Capitano Nicolb Dubrocovich,for whom I borea letterfrom the friendly Prof. Glavinic. He led me to his house,introduced me to his family,and made me feel thoroughlyat home. Most of the mayorsin the Dalmatian Archipelagoare retired Capitainesde longcours, substantial men with large estates,who, duringtheir voyages,have accumulatednot onlycapital, but a large stockof refinementand generalinformation. In this point insularDal- matia much resemblesSwitzerland. Gelsa, in the officialmap Gjelsa, and by the Slavs called Jel?a, is oftenmentioned in local history. It appears to have been an ancient city,whose name is now lost. The Statutodi Lesina, compiledin A.D. 1331,t as the prefacesays, by a Russian lawyer,whom lung-diseasedrove to Venice, tbus speaks of the ruins in A.D. 1407, Civitas Vetus (Gradina) in Jelha(p. 202); and in 1425 we read (p. 217) "inter civitatem Jelsee.... prope civitatemveterem in Jelsa." As will be seen, it was presentlydistinguished from the otherCivitas Vetus (Cittavec- chia) by taking the name of its fine fountain. Also Rafaele Luacovich (in Farlati Ill. Sacr. i. 197) says," Pagos habet (the island) civitatibushaud absimiles; Gelsam fontibusperennibus divitem. Verboskamubertate agri pinguem,"&c., &c. The townletoccupies the head of a bightopening to the north- east, crossed by a masonrydam, and made tolerablysafe for shipping by two short moles on the northand south sides. A few good houses, amongstwhich is the Podesta's, preferthe northernshore; the mass of the settlementlies opposite. It evidentlyowes its origin,as well as its name, to a perennial spring of the purest water,which wells up near the piazza. oil compelledhim to removethe " Fabbrica di conserve" to Barcola (S. Bartolo), in the Bay of Trieste. * All thenames of towns(except Cittavecchia), hills, and dales are purelySlav. The orthographywould be Vrbanj and Vrboska. t Of this Liber StatutorumCommunitatis Pharce, afterwards printed under the name" StatutaCommunitatis Lesime," Venetiis, 1643, see Boglic (pp. 112-129).

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Pharia and Gelsa di Le8ina. 291

The Statuto (pp. 50, 51) says, " usque ad fontemvocatum Jel- sam, quee est apuclmare salsum." It has its " Parrochiale"- Half churchof God, half castle 'gainst the Moor; ;and its cemeteryaround the desertedAugustine monastery. It wears a thriving look, and I heard with pleasure the sound of the bell whichrang the boys to church. A wireconnects it with Spalato, the chef-lieu;and, indeed,nowhere about the Mediter- raneanhave I found this instrumentof our later civilisationso generallyused as in Dalmatia. The principalwant here is com- munication; thosewho would embarkmust ride forhalf a day over the roughest path, spanning mountainsand valleys, to Lesina town,the only stationof the AustrianLloyds. In 1772 Fortis (ii. 178, 179; Fr. ii. 242) describedGelsa as "un grosso villaggio ben situato,"well-peopled with citizens dresseda la Francaise, and boastingnot a fewgood houses and villas. The port receivesmany perennial streams, and thehills, which slope gentlyto the sea, produce the finestmarmorine breccias,which are used forcoarse pavement, or are built up in pauper huts. Besides the " corallata" beforementioned, there is a stonewith irregular stains of pavonine colour,taking a polish which equals that of the finestRoman conglomerates. Mgr. Blascovich,Bishop of MAakarska,took fromits quarriesall the material for the columnsof his new cathedral,and forthe steps and pillars of his altars. Unhappily,the cutterswill save timeand troubleby contentingthemselves with the upperstrata; the softerportion, especially the natural cement,containing the marble,is thusdegraded by theatmosphere, perhaps also by sea- water,and it soon deteriorateswhen exposed to sun and rain. Fortis also remarkeda black and whitelumachella, composed of hardenedbituminous earth, and orthoceratites,changed, as usual, intoa sparrysaline calcaire. He does not noticethe ruinseither at Cittavecchiaor at Gelsa, for the good Abbe was a naturalist ratherthan an antiquary,and he seems somewhat sore about his reception. The fishermenwere perhaps at sea, and " quando io giunsi cola non trovaiquella cortesia,che suole abitarecolla povera gente." Thus his experienceand mine differtoto celo. Tuesday (December29th) appeared in its veryugliest guise. The climate of fair Dalmatia can, like certain Madonna-faced beauties,show an amountof ill-temperas seriousas it is startling. I rose in the dark,hardly hoping to see my host,but he was no "marinero di acqua dolce," justifyingthe old saw, Di sera leone, Di mattinababbione; and he cared not a jot forthe frenzied gusts,the Scotch mists,

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 292 Capt. R. F. BuRToN.-Oiz theRBined C0iieS of and the showerswhich fell as if bucketswere being emptied upon the hiUls-in these latitudes,when it rains it doesrain. During the evemnigwe had learned that two ruinswere to be visited,the "Grad " and the " Tor," the formerbearing from his house, "Ostro quarta e mezza Sirocco" (S. by i E.), and the latter" Ostroe tre quartini Levante " (S. E). The time would take an " oretta"-beware of the " little hour" in Dalmatia and Istria. On this occasion,however, it was only double. Passing throughthe townlet,we began the ascent of the low and ruddyoutliers of the Gvezdjena Gora (Gvezd Berg of the map), or Iron Mo3ntain,a long range runnimgnearly upon a parallel of latitude. The reason of the metallic name is unknown; perhaps it is derived from the steel-greYspines, bands, and cornices of the hardest limestone,which accident the slopes. Beyond the Madonna della Salute the goat-path became stiffand stony,slippery withal undermud and rain. On these islands the travellerin searchof prehistoricremains becomes,after a fashion,an explorer. He mustvi'sit everythin{ that bears the name of " grad," or its multiformderivations, and, as in Africa,he must labour to ascertainwhat thereis not, as well as whatthere is-ea queesunt, tanquam ea quamnon sunt. Afterbreasting the iron height, we reached this particular " grad," and the firstglance told me that the masonry,which might have been Venetian,was more probablypost-Venetian and Slav. The aneroidat 28 9, and at a sea-level,29 9, showedan altitudeof a thousandfeet, and the site was thatof an Etruscan city, a " Mull," the Icelandic " Mdli," or loop of high ground, with a declivitymore or less precipitouson all sides but one- here the western. The rock fell sheerto the south; the neck had been fortified,but the outworkswere so ruinousthat their formcould hardlybe aseertained. The enceinte followed the contourof the ground; in places it was based upon the limestone, at this and in other parts formingparallelopipedons by stratificationand cleavage, which easily suggested the " Cyclopeanwall." Here and thereit is difficultto distinguishthe natural fromthe artificial,so exactly do the joints correspond. Fortis remarksthe same of the sandstone,and he figures(Table xii. vol. ii. 100) the naturalwall (" filonisimili a muraglie") of Rogosnizza. In Istria I should have thought that the origin of this " grad" mighthave been one of the so-called Castellieri. A dilapidatedcistern-shaped affair, showing the spring of the arched roof,and facingsouthwards with westimg,is, according * The curiousreader will consultthe paperon the "Castellieriof Istria," whichit was my evil fateto publishwith the AnthropologicalSociety of London('Anthropologia," No. iii.Oct. 1874).

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 3ourn..Aaitf,rpobo0 ItyIot87YlXI

E LE VATI1O N O F T HE TOR. Shos t -dutn bos.s

0.3l/ OT68 1713r 17 04

07 l, 70- 1-1 rt0 07",4 .2 T . E a93 iswl. .. t i . o N _ z

90 > 7 -t, ] NX 27 O!n"o'2 ,

44E~~~~~~~~3C) F 'K . M5

DALMAT I A N F IN 0.

\> r j 2...... I ...... A...L...M...AT'A2.I N2lS I~~~~~~~ J a ~~~~~~~~~~~~~~~IN

BoFz,af~~77z14 f r Og r

I-C.F. -i&,J eL Mao E. C.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Pharia and Gelsa di Lesina. 293 to the legend,the burial-placeof its "king," Subich (Subij). All tracesof the tombhave lately disappeared. ProfessorBoglic tells us (pp. 126, 127) thatthis " Grad Gali6- nik " is describedby the monkPriboevo, of Lesina, and quotesas followsfrom a manuscriptletter addressed to H. E. Gio. Bra- gadin in A.D. 1723: " Sulla di lui sommita(the mountainnear Gelsa) v'e un recintodi mura,da Oriente (poggia) sull 'orlo di sasso dirupato. L'opera e in quadro,lunga 30 e piu brazza (a minimumof 21-83 yards); larga nell' ingresso brazza dodici (8&73 yards); in fondo brazza otto (5-82 yards) incirca. Si vedononel di lui mezzo due casette,una diroccatama angusta; l'altra a volta di pietra, bastante a ricettaresotto il tetto la staturadel piiu alto homo,ma capace di poche persone..... Intornonelle murainteriormente si distinguonocontinuati buchi di travi (the modernbuchi per i falconi,dowel-holes to receive the ends of floor-joints)di mezzo carroin linea uguale. Danno a crederesi sia statoun tavolatointorno per gente,che coperta da merli delle mura possa tutelare il recinto... . Questo luogo vien chiamato Galicinich. Corre un idiotismo (trivial legend), che due regoli fosseropatroni, di Gradina, che nell' idioma illirico suona citta diroccata,e di Galicinich; tra loro inimici,et in continuaguerra; e vi fu chi in Galicinichegual- mentecredulo ad avaro sudasseper trovaretesoro nascosto." Somewhatdespairing about the otherbuilding, I walkedacross the slope of the Iron Mountain,about three-quartersof a mile to the west,by a path roundingthe heads of two small ravines. At one shelteredspot appeared a newly-plantedvineyard: in Dalmatia, as in Istria, cultivationrises high above sea-level. Presentlywe came upon the Torre di Gelsa; the Slavs call it Tor,"" or sheep-fold(e.g. ui-Tor,nell' ovile), but perhaps the latteris a merecorruption of the former. I was delighted; my rough and rainywalks had not been in vain. The site is singu- lar; the apex of a rocky arete, utterlywithout water, except fromrain, and apparentlyisolated, although large cut-stones, whichmay have belongedto it or to its outworks,were scattered around. The inside was filled up with earth; externally it showedfrom four to fiveisodomic lower coursesof large ashlar, calcaire fromthe mountainon which it stood,and nowherewas therea trace ofmortar. The largestparallelopipedon measured 2-06 metres(=6 feet9 inches) by 0-76 (==2 feet*59 inches) in height. The angles, especially the north-eastern,showed the draught extendingthrough the coursesfrom the lowestto the highest. The westernexterior consisted of fourlower coursesof large stones,capped by threemodern, or, at least,smaller layers; and the emplectori,or "old English bond," popularly called "headers and stretchers,"were apparentlynot unknownto the

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 294 Capt. R. F. BURTON.-Oll theRuined Citiesof builders. The stoneswere all boldly bossed,like those of Salona, with chiselleddraughts, and the heightof the projectionmight have been 6 to 8 inches. The magneticmeridian passed through the angles, and the slope was an oblong ratherthan a square. The northernside measured 7-25 metres(=23 feet 9 inches); the southern7-14; the eastern and the western6-66 (21 feet 10 inches). My sketchwas utterlyspoilt by the delugingrain at the time of inspection; the nebulaemalusque Jupiterdetermined to do theirvery worst. But my excellent host promisedto send me plan, elevation,and measurements.He kepthis wordwith truly Britishpunctuality, and enabled me to presentto myreaders the sketch which accompaniesthese pages (P1. xiii. figs.3, 4). I gladlytake the earliestopportunity of thankinghim once more. Professionalarcheologists will determinethe ra.pportbetween this mysteriousbuilding and the Pantellarian "Sesi," which have yieldedstone implements; the Nuraghi of Sardinia and the Balearic Talajot (Arab. .,l%, i.e. watch-towers). The general aspect remindedme of the garrison-stationson the Roman high- roads,especially of that near Khan Khuldeh,supposed to be the " MutatioHeldua" of the JerusalemItinerary, near Bayru't,on the way to Sidon. But here there is no sign of cement. I fotud no traces of a highway,and the site, commandedon the southernside, and occupyingthe roughest of rockyground, where enemiesmight everywhere lie in ambush,and where halfa dozen square yards of tolerablylevel surfacecannot be had, renders it equally unfitfor a refugeplace and fora settlement.It might perhapsbe an outworkand a look-outcommanding the sea; still there remainsthe curious contrastof elaborate finishwith an object forwhich the simplestbuilding would suffice. Prof. Boglic appears to think(pp. 11-12) that the " Tor" was connectedwith certain ruins near the Gelsa townlet,which have disappearedonly lately, either buried under alluvium or removed by the peasantry. He quotes Priboevo of Lesina, who,in a dis- course or harangue (De Origine et SuccessibusSlavorum) pro- no&unced(1525) before an "Academy" in his native town, describesin thesewords the ruinsof Cittavecchiaand Gelsa. I preservethe quaint contractionsof the original. " Quapp. nemine capiat admiratio, duas olim, altera ab oriente,altera ab occidente,cui' supra memin' egregios sortita port',hoc i agro urbisextitisse, ipsa ruinapristina ear. dignitate prodente, apparentibus ibidem pluribus magnis mdificiis,et ex pario lapide truncatis,et semicorrosisHereum imagiib', nec non et lithostratisvarias bestiar'et sydir'formas preferentibus. . . . Est et unum adhuc fereintegrum inter montis oppidum urbi (i.e. Gelse) quea orientalemagri hui' regionemobtinebat

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Pharia and Gelsa di Lesina. 295 superemimens,juxta quod ad jactum-sagitte, ex qnadratismnire magnihtdinislapidibus turnis est, unicurn habens angustum hostio- iun, in qua ob antiquitates nullurn cemenitivestigitum apparel. Sunt et in plerisq. aliis hujus insulwlocis, humanaehabitationis vestigia,in quib' antiquiilli Pharensesvicatim habitaverunt. Ex his colligo pter primariain qua sum' urbem,que a meridionali insulkparte posita,Pharum propie, vulgo autem(sicut Calepinus prodit) Lisna, corrupti'aut Lesina dicitur, sex in hac insula olim oppida absque villis, quee numerosxeerant, extetissi" (sic). Prof. Boglic (pp. 11, 12) thus describes the building: "Anche una torre,che sorge da un cumulo di sassi sul monte posto a mezzogiornodi Jelsa,offre tutti i caratteridei ruderidi Cittavecchia,soltanto i massi sono battutipiti rozzament',ed il lavoro potrebbecredersi pRt antico. Alcuni di questi sono alti due piedi ed &to polici Viennesi (== 2 feet9'10 inches); lunghi cinque (- 5 ft. 2 23). Ai tempidel Priboevonel lato di mezzod'i vi era una porticina,che poscia ruino,e non restanoche soli tre lati, che varebbe la pena di meglio conservare. I pastori ne distrusserouna parte,atterrando poco a poco le pietre, che sovrapostele une alle altre senza cimento,piu' facilmentesi potevanosmuovere. I lati della torreche ancora restano,sono alti sedici piedi Viennesi circa (== 16 feet7-12 inches); lunghi a Settentrioneventiduie (= 22 ft.9 79) diciannove(=19 ft.8A46) ad Occidentee Levante." ProfessorBoglid wisely refusesto "evocate" the Pelasgi, because similar monumentsare found in Eastern Asia, for instance,where those wandering tribes cannot,"' even with the greatest indulgence," be made to emigrate. I would here remarkthat the SamothracianPelasgi seem of late yearsto have gone out of fashion,probably on account of the highly absurd etymologiesproposed for them, such as Semitic roots for an Aryan race; * and yet thereis no race of which the universal voice of classical antiquity speaks with more clearness and consent. The professorholds thatthe Gelsa styleof architecture was not peculiar to a single people, but ratherthat it denotes a certain stage of civilisation,of progress,evolved under ethno- logical conditionsthe mostdissimilar. A nationwhich either cannot,or whichknows not how to make the cement that con-

* For instanice,Pelishti or Felishti,peregrinus, advena, from Palasha (Falasha) iuigravit,whence, by-the-by, the Falasha Jewsof Abyssinia. Some explain the Semitismby supposingthat it was applied to the Indo-Europeansby the Egypto- Pheniicians;if so, had theyno racial name of theirown? Also fromPelasgus, the kingwho civilisedthem. From Pelargos,a crane (quod gregatimerrarent)l from Palueos,ancient; from Pelagus, the sea, beinga maritimerace; from the root of "pellere" and ex-pel. Chabas (":etudes sure l'antiquite,"'&c., August, 1872) thinksthat he findsthem in the Pelestas of the Stela of Medinat Habu, recountingthe conquestsof Rameses,and datingfrom the 12th centuryB.C.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 296 Capt. R. F. BURTON.- On theRuined Citiesqf solidates small stones into a mass of masonry,must perforce, when building its defences,employ megaliths whose weight keeps themin place, evenwhen exposed to the shockof battering engines. He findsthis speciallydenoting, in Greece as well as in Italy, the epoch of incipient settled life, the first stage of civilisation. He thereforeconcludes that the ruins of Citta- vecchia,and of the Tor di Gelsa, were the worksof the later Illyrians.* These tribeswere found occupying Lesina in the 4th centuryB.C., and in B.C.45, according to Diodorus Siculus (xiii. 3), they inhabited "a little settlementof extraordinary strength(&v rtvt Gi)cpta wao' v7rep3o&XvjXvp5). I musthere join issuewith the learnedprofessor upon sundry points. Diodorus Siculus (xiii. 3) speaks distinctly of the founding (1criGtls)and walling of Pharia by the Parians. The walls of Cittavecchiacertainly belonged to a Greek colony,as is proved by the thousand remains foundwithin them, to say nothingof theirshape, and the formwhich they enclosed. The Tor cli Gelsa does not appear to me a ruderor a moreprimitive formthan the defencesof Pharia, althoughhaving beenexposed forlong ages to the violentBora, it has been moremutilated by time and weather. And I can hardlyimagine how he made away with the fourthside; it remindsme of what was said of Ireland, where- The sly surveyorsstole a shire. Finally, the remarkable resemblance,amounting almost to identityof shape, between the two Lesina ruins and those of the " Murazzo," or long wall of Salona, the subject to which the firstpart of this paper was devoted,naturally suggests that all threewere the work of a single people, and that people not the barbarousIllyrians, but the comparativelycivilised Greeks. A fewwords to the reader by way of conclusion. The late Mr. Paton, whose acquaintanceI had made,and whose memory I would treatas that of a friend,wrote a useful and not un- learnedbook, which he called by the picturesquename, " High- lands and Islands of Dalmatia." I should be sorryif it induced the archaeologistto imaginefor a momentthat explorationof eitherfeature had ever even been attempted. Whilst Greek and Roman antiquities have occupied the lives of many able men,prehistoric study is only now making itself known by name. The best proofis that at the end of 1874 onlytwo stone implements,both found within the year,had takentheir place in the littlemuseum of progressiveSpalato (P1. xiii. figs.1, 2). * Prof. Boglio (pp. 9, 10) adopts the opinionof Niebuhr, that-(1) the Illy- rians were a differentrace from the Liburnians; and that (2) the Illyrials, fromwhom the Skipitar (Albanians) are descenlded,inhabited with the Greeks the Pceonians,and the Thracians, "Pelasgia," or Macedolnia; but they were neitherGreeks nor Pelasgi.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Pharia andGel.a di Lesina. 297

I need not speak of the Dalmatian highlands; but to explore the islands there is some little difficulty.The travellerwill findscanty aid fromsteamers; he will have no roads,and he must work his way on foot through the roughest bush; he must carefullyvisit everyheight; and he mustbe preparedfor few successesand many failures. If he cannot speak Slav, he mustbe accompaniedby one who does, and he had better take with him plain and coloured engravingsof stoneimplements, whichwill supplythe want of technicallanguage; forinstance, an arrow-heador axe will be knownonly as Mlalikamen -strella, the little lightning-stone. He must be prepared to rough it, to live hard, not to murmuragainst the smaller insects,and, perhaps,to risk an attack of ague and fever. Finally, though life is not expensive,he must prepare fora considerablewaste of precioustime. I had preserved,so to speak, the Dalmatian Islands as a happy hunting-ground;but officialoccupations, and, worse still, a serious illness, interferedwith my projects. Next to exploringfor oneself,the best work an explorercan do is to promoteexploration in others. My highest ambitionfor these pages is to show how much remains to be done. A partyof threeor four friends,forming a committeeof discovery,could hardlyspend theirtime better than by devotingthe best season of -theyear, from April to June included,to a carefulsurvey of the Dalmatian Archipelago,visiting every site called Grad,and collecting the folk-lorewhich everywhereabounds. I prefer the numberfour, because it would obviatethe delay by enabling the partyto separate into two and threesections. Needless to say that all my small amountof experiencewould be gladly placed at the adventurers'disposal, and that they would have my best wishesfor their success.

EXPLANATIONOF PLATES XII. AND XIII. Plate xii.-Plan of Salona, near Spalato, showingthe Long Wall; withplan of the Amphitheatreand CaosarianGate. Plate xiii.-Fig. 1. The first stone implementfound in Dalmatia, now in the Museum of Spalato. Half natural size. Fig. 2. Axe of greenstone,found at Salona. Half natural size. Fig. 3. Elevation of the Tor di Gelsa, in the Island of Lesina. Fig. 4. Plan of the Tor di Gelsa. DIscussIol. Mr. HYDE CLARKE said thatCaptain Burton's paper was of the greaterinterest, as he had observedthat in Istria and Dalmatiathe ancientnames were largely pre-Hellenic, ancd conformed to thosein Etruria,Hellas, Asia Minor,and generallyto the Sumerianclass.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 298 Disceussion.

Prof. HUGHES said that Captain Burton had broughtbefore them the resultsof his observationson threequite unconnectedremains in differentparts of the country. 1st, there were the neolithic remainson the island; 2nd,the ancientcamp; and 3rd,the old walls. With regardto Captain Burton's remarkson the first,he- thoughtthat, whatever reason there mightbe for supposingthat great changeshad taken place in the physicalgeography of the MIediterraneanarea sincethec diminutive Maltese elephantand its contemporarieshad walkedfrom Africa to Malta and Europe,or- thatPaltolithic man had looked on a shorevery different from what we see therenow, still we haveno evidencethat would justifyour assumingthat the Adriatic islands had been cut off from the mainland sinceneolithic times, even though we knowthat along the linesof volcanicactivity considerable changes of levelhave takenplace in historictimes. With regard to the second set of phenomena,it was interestingto findthat in that partof Europe,as in Britain, the raceswhich immediately preceded historic times built the same kindof camp,occupying a hill-topand throwingup irregularlines, wherenecessary, and where the position could be most easily defended,but whichhad no constantsize or form. Pottery,however,he thought,was an uncertaintest of age -whenapplied to the correlationof remainsfound in districtsfar apart and countries imperfectlyworked out. For instance,he had seen potterybeing manufacturedon the northernspurs of the Pyreneeswhich differed in no respectfrom that found in caveswith the remains of primeval manin the samedistrict. He objectedto Captain Burton'sdefini- tionof Cyclopeanand Pelasgic,pointing out thatthe termPelasgie had got into disreputeamong ethnologistsbecause, if appliedto everythingpre-Hellenic, it was too vague for theirpurpose, and when limited,the limitationshad been generallyfounded upon unwarrantableassumptions. Cyclopeanwas used forthose ancient walls,built of stones so enormousthat fancy called up giantsfor their construction.But there was not sufficientevidence to fix the relativeage of Pelasgic and Cyclopearr,or to justifysuch a new applicationof old wordsas thatsuggested. Mr. MOoGGRIDGEsuggested whether the objectof whicha model was exhibitedmay not have been a Muller. It closelyresembles a, Muller whichhe found in disinterringone of the most ancient churchesin this island,that had long been hiddenfrom the eye of iilan; -solong, indeed, that when he wentto workthe onlyclue lhehad was thattradition said, c underthat sand-hill lies theancient church." If this be a Muller,it mayindicate the nationalityof thosewho at one timeoccupied the spot on whichit was found. ThePRESIDENT remarked that as CaptainBurton had alluded to the- resemblancebetween the Castellieri he had spokenof and someof the hillforts, described by him(the President),in the " Archaologia,"- it mightbe useful to say a fewwords on that point. It was cer- taiiilyremarkable, but by no means surprising,that sucha resem- blanceshould exist, the chieffeature of whichappeared to him to consistin the fact of theirbeino situatedon the tops of hills,.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions Discelssion. 299

remotefrom water, security against attack being the main object sought,and imDlyinga conditionof constantwarfare between tribe and tribe. A thicklyovergrown, marshy, and unhealthylowland countrywould also have this result. Whereas, in times correspondingto the Romanera, the encampmentsand otherhabitations were oftensituated in low ground, and generallynear water, showingthat in a highercondition of culturethe interioreconomy, comfort,and convenienceof the inhabitantswere beginningto be betterlooked after. The casts of spear-headsexhibited by Captain B3urton,the originalsof whichare said to be of copper,are of the kind more usuallyconstructed of bronze,and are of whatmay be called the northerntype; thatis to say,they are leaf-shaped,with a medialrib, and providedwith sockets, the wholeof whichis cast; correspondingin all respectsto those foundin England,Denmark, France, Germany,and eastwardas far as Siberia,and differing essentiallyfrom those found in the islandof Cyprus,from those discoveredon the site of Troy by Dr. Schlieman,from some of thoseused in ancientEgypt, and fromthose found in Hindustan, in whichthe spear-headsare eitherfitted into the shaftsby means of tangs on the sockets,or formedby being cast flat,and after- wardsbent over the head of the shaft,leaving a longitudinalslit at the pointof junction. This is a noteworthydistinction between the implementsof the two regions. The fragmentsof pottery exhibitedappear to be of at least two kinds; a finedescription, correspondingto whatwe shouldin thiscountry call Roman; and a somewhatcoarser kind, containing grains of quartz, resembling whatwe mighthere term Romano-British; but thereis none of the looser, ill-baked kind, which is characteristicof early British pottery.Captain Burton had been so kilndas tomake him a present of thesecasts for his collection at BethualGreen, acnd he need hardly say thathe regardedthem as of the highestvalue and interest. The authorbriefly replied.

M1r. Rudler read a communicationfrom Mr. HORACE B. WOODWARD, describinga wooden image and a spear-headdis- coverednear NewtonAbbot. They were foundby workmenin the employof MessrsWatts, Blake, Bearne, and Co., clay mer- chants,of Newton Abbot,by whichfirm they were exhibited. In digging for the pipe and potters'.clay belonging to the Bovey formation,and which occupies the greater part of the valley between Bovey Tracey and Newton Abbot, there is always met with a superficialaccumulation of gravel, sand, and mud, called the " Head " (and sometimes" Pengelly's Head"). This gravellydeposit has no connectionwith the Bovey deposit, being comparativelyof very recent date; nevertheless,when looked at in a large wvay,and tracedon the ground,it is found to extendup the hill-sidesat Woolborough and Milber Down, to a heightof 330 and 420 feet,and, indeed,in one place, to VOL. V. Y

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions 300 H. B. WOODWARD.-OnImage, etc., from Devon.

a height of 540 feet. At the same time, the gravel extends towardsChudleigh, and occurs in outlyingmasses, which seem to connectit withthe driftson HalclonHills, 800 feetand more above sea-level. It is necessaryto mentionthese facts,because the valley of the Teign, near Newton,has received deposits of driftgravel whichcannot be satisfactorilyattributed to present riveraction, and may,perhaps, be connectedwith the phenomena of the glacial period. Be this as it may,there is no doubt that some of the gravels in the lower parts of the valley have been re-assortedby the presentriver; so that,although it may be difficult,and, in some instances,impossible, from superficial observation, to separate the gravelsinto those of most recentdate and those of possibly glacial age, yet we must be prepared to assign a very recent originto manyof the gravelsin the low grounds,and to account for rapid accumulationof gravel by the fact that it was there " ready-made,"and had,only to be re-assortedand re-deposited by the action of the presentriver, aided as well by the deposits broughtdown fromthe hills by the main streamand its tribu- taries, as by mud that might be broughtup by theincoming tide,for tidal conditionsformerly prevailed much higherthat at present. The specimenswere discoveredin the gravel above the clay betweenNewton Abbot and Kingsteignton,in a large pit by the river-sideon its leftbank, about a quarterof a milenorth-west of theturnpike. The clay pit is called ZitherixonPit, and is partly in the parish of Teigngraceand partlyin that of Kingsteignton. The section displayed a thicknessof 25 feetof gravel,sand, and mud stratified,but varyingmuch in detail at differentparts of the pit. Towardsthe top the gravel was generallyfine, and at the base usually verycoarse. Many fragmentsof oak trees were met with-the heads pointing down-stream-and against themgravel was banked on the higher ends (up-stream),while mud was accumulatedon the lower end. The oaken figurewas foundstanding in a nearlyupright position against one of these trees,embedded in gravel,and at a depthof 20 feet. The bronze spear-headwas foundat a depth of 15 feet,and numerousbones of ox, deer, &c., were met with at a depth of from15 to 20 feet. The image and spear-head have since been described and figuredby Mr. Pengelly in the " Transactionsof the Devonshire Association,"vii. p. 200, and plate. The level of this depositwas but 4 or 5 feet above the river- level,so thatthe pit wasworked to someconsiderable extent below the water-level,and much subject to infiltrationof the water. The depositmay, without doubt, be classed amongstthe latest accumulationsof the river,and be included with the alluvium.

This content downloaded from 195.34.79.223 on Sun, 15 Jun 2014 22:28:14 PM All use subject to JSTOR Terms and Conditions