Arthur Evans.

Antiquarian researches in Illyricum : Part I-IV Communicated to the Society of antiquaries online

. (page 1 of 32)
Online LibraryArthur EvansAntiquarian researches in Illyricum : Part I-IV Communicated to the Society of antiquaries → online text (page 1 of 32)
Font size
QR-code for this ebook

/ i^v'tV (^i [ i /»/«-^

/ *

v^^?^ ^/.















Antiquities of Epitaurum.


4. The original site of Epitaurum Eagusa Vecchia and not Prevlaka as suggested by Jlommscn.

5. Greek coins and gems found on the site of Epitaurum.
8. Existing architectural remains: the Aqueduct.

1 1 . Bath-chamber or Piscina at the head of the Aqueduct.

12. Monument to P. Corn. Dolabella.

13. New Inscriptions, one mentioning ' ^dile ' and ' iivir Quinquennalis.'

16. Development of Civic Institutions at Epitaurum, as illustrated by monuments.

17. Gems relating to cult of ^sculapius: this cult apparently extinguished here by St. Hilarion.
19. Discovery of Mithraic monuments near Epitaurum.

22. Observations on some Mithraic gems.

26. Engraved Christian gem, probably representing Vision of Constantine.

27. Roman Christian ring.

27. Observations as to the date of the destruction of Epitaurum.

Antiquities of the District of Canali.

29. Derivation of the name by Constantine Porphyrogenitus explained.

31. Illyro-Roman survival in tlie local nomenclature and physical types.

36. Apparent site of lloman Municipium at Sveti Ivan and Djare.

37. Monument mentioning the 'iivir ivre dicvndo.'

39. Traces of Roman road leading from Epitaurum to Risinium.


Antiquities of Rhizon oh Risinium.


4(1. Heinains ol Acropolis at Risano.

41. The '..Eaciaii ' walls oC the ancient city-

4"J. JUviian coins struck at Risinium.

44. l!reek terra-cotta vase and Askos trom this site.

45. Notes on the Greek commercial connexion with the lUyiuin coast.

46. Roman inscriptions.

48. Traces of Aqueduct and Reservoir.

49. Christian intaglio.

4!». The Kisinian episcopate in tlie sixth century.

,')(>. T.atc Roman i-namelled pendant displaying Persian influences.







iv— ^A. v\\";#e»«.. ;■; \ ■^■- V'-'ir^-^'




s» ^ ^^ -J V. '*•.": ltd

'tVmr'iU- %.-. :. (



PROM9NAY ■ ^ -

■. f:-i.,l.,f





■'"•■>" ■ POL.

■ (', .

■A-' o\

3,A(iVR I V Mjj^_^^i^j~^


J»'/;r <^ 'Kcman- IStuUdpiO'

have been (hsci^r4-ed/ )

Doiibtfi.ll Reman, Sites \

and J\'ame.s J

SiTTUZn RoatLs -

I'cnjecticfal CiTurse- of \ .

Roma/i roads J


Q i ip 22 JO





Jiijiicutijuf tJie (cia'.se of the- Hoods and the' Sites
where Honuin RemajJis /uwe been/ discovered-.

Prepared/ hy t/ia Author.

C FKeaLith S.Castlo Si Holbam Londoni C






r* T 1



Owing to the neit^hbourhood of the civilized republic of Eagusa, Avhich sprang
as it were from the ashes of the Grseco-Eoman city, the antiquities of the Dal-
matian Epidaurus have been investigated from the early days of the Renascence.
The merchant antiquary, Cyriac of Ancona, who visited Ragusa during his voyage
into the Levant, undertaken in 1435, had already begun the work of copying tlie
remaining inscriptions, which was continued in the next century by the native
Ragusan antiquaries, who supplied Aldus Manutius and others with epigraphic
materials from the Epidaurian site. The work thus early begun was worthily
continued in the last century by the Ragusan patrician De Sorgo," more recently
by Dr. J. A. Kasnacid and others, and Professor Mommsen personally collated
many of the inscriptions for the great work of the Berlin Academy.'' The aque-
duct and general antiquities of the site are treated at length by Appendini, but
in a somewhat fantastic and uncritical manner.'' A residence on the spot has now

" Comment. Lud. Cervarii 2'uheronis de origine et incremento Urbis liliacusuncc. Eagusa, 1790.

^ The hitherto known inscriptions from t]ie site are collected in C. I. L. iii. p. 288 seqq. and Prof.
Mommsen (.?. v. epidaurum) gives a resume of the earlier sources for the epigraphy of the place.

"^ Notizie istorico-criiiche sulle Antichitd, Storia e Letteratura di Ragusei. Ragusa, 1802, t. i. lih.
i. ii. The remains at Eagusa Vecchia have been touched on since Appcndini's time by Stieglitz, Istricn

H 2

4 ^Lntiquarian Itescarchcs in lUi/ricum.

enabli'd luc fo make some fresh conti-ibutions to the materials ah-eady collected,
and to correct perhaps some prevailing misconceptions.

The site of the ancient city, at present occupied by a small town called, by a
curious transference of names, Ragusa Vecchia, l)ut still knoAvn to its Slavonic-
speaking inha])itants as Zavtat or Cavtat, from the earlier llomance form Cicifofe,
is on a small pciiiiisiila jutting out from the opposite side of the bay to that on
wliieh its oifsin-ing Ragusa stands. Although tlie Dahnatian Epidauros, or, to
accept the prevalent local orthography, Epitaurum," does not appear in history
till the time of the Civil Wars, the name itself mav be taken as a sufficient
indication that it was an Adi'iatic colonial station of one or other of its Pelo-
ponnesian namesakes; and its peninsular site Avas just one of those which offered
special advantages to the early Greek settlers on a barbarian coast.

Mommsen, indeed, who visited tliis site in order to collate^ the monuments for
the Corpus Inscriptionum, has revived in a new form a theory, already propounded
by Mannert,'" and others, that the site of Epitanrum is to be sought at Prevlaka,
at the entrance of the Bocclie di Cattaro, and not on the peninsula of Ragusa
Vecchia. It has been pointed out by these authorities that the Tabula Feutln-
geriana makes Epitaurum 105 miles distant from Lissus and 103 " from Narona,
while Pliny '^ makes it equidistant — 100 miles from citlicr — and it lias been urged
that these measurements can only be reconciled with the position of Prevlaka.

As Mommsen however himself admits, the statement of the Itinerarium
MarUimuDi " that Epitaurum was 200 stadia from the isle of Melita (Meleda) can

und Dalmazien, p. 2G4 (Stuttgart unJ Tubingen, 1845), Wilkinson, Dalmatia i. 373 (London, 1848),
Kohl, Reise nach Isirien. Dalmazien und Montenegro, ii. 33 ser/r/. (Diesden, 1856), Lago, Memovie sulla
Dalmazia (Venezia, 1870), and others, but the notices are slight and add little to our knowledge.

' On a Privilejium Veteranoru7n of Vespasian found at Salona there is mention of a P. Vibius
Maximus, — eimtavu . kq . r. In the Tabula Peutingeriana the name appears as Ej)ilau7-o: in the Geo-
grapher of Ravenna as Epitauron (."79, 14) and Epilnunnn (208, 10). In St. Jerome {Vita S. Hilarionis)
Epitaurum : in the sixth century Council-Acts of Salona, Epitaurensis Ecclesia. The town is alluded to
by Constantine Porphyrogcnitus {l)e Adm. Imp. c. 29) as tu Kdarpov ro imXeyvfitvov IliVaupa; and its early
Slavonic name was Starigrad Pitaxir, still preserving the t in preference to d. The readings of Ptolemy
(2, IG, 5), Pliny (23, 143), and Antonine {It. Mar. 520), cannot weigh against this consensus of local
testimony; but we need not with Prof. Tomaschek {Die vorslaivische Topographic. &c. p. 37) seek an
Iliyrian derivation for the name.

" 7, 350.

' Accepting the correction of the distance Naron,-; — Ad Turres (see p. 79).

J Hist. Nat. iii. 22, 143.

<^ .4 MELiT.t EPiDAVRos STADIA cc. /(. Antonini. 520.

Antiquarian Researches in Illyricum. 5

only be reconciled with the K-agusa-Vecchian site. He further observes that
any one Avho, like himself, has visited Ragusa Vccchia, who has seen the remains
of the amphitheatre cut out of the solid rock, the traces of the Homan harbour,
the inscriptions which, though not presenting in a single case the name of the
city, are numerous and imposing, and the other abundant traces of Roman
habitation that are daily brought to light, can fail to recognise the fact that
a famous and important Roman city must have existed at this spot, epithets
which, among all the Roman stations on the coast between Lissus and Karona,
alone apply to the Colony of Ejiitaurum."

In order to reconcile these conflicting indications Mommsen has recourse to
the hypothesis that the original Epitaurima existed at Prevlaka, but that for some
reason unknown, and at a still flourishing period of the Roman Emjoire, it was
transferred to the Ragusa-Vecchian site ; so that there woixld be an Old and New
Epitaurum as well as an Old and New Ragusa.

This hypothesis, not very hopeful in itself, appears to me to be untenable for
several reasons. At Prevlaka a single inscription only has been discovered, refer-
ring to a decurion of the Sergian tribe,'' the tribe to which the citizens of Risi-
nium and the Roman predecessor of Cattaro belonged, but not the tribe of the
Epidauritans, which was the Tromentine . Taken by itself, therefore, this inscrij)-
tion supplies internal evidence that it belonged to one of the known Roman cities
of the Rhizonic Gidf . A careful examination of the isthmus and peninsula of
Prevlaka has convinced me that no ancient town has ever existed at that spot."
Not only are all architectural traces wanting, but the soil is absolutely deficient
in those minor relics, such as fragments of pottery and tiles, that always mark an
ancient site.

On the other hand, tlicre liave been discovered on the site of Ragvisa Vccchia ,
indubitable relics of Hellenic intercourse, dating from prae- Roman times. ( /

'^ C. I. L. iii. p. 287, s. v. epidaurum. I do not know to what Prof. Mommsen refers as the remains
of tlie Amphitheatre.

" C. I. L. iii. 1738.

° Dr. Ljubic, Viestnik hrvatslcoga archeologickoga Druztva {Jotirnat of the Croatian Arcliceological
Society), iii. p. 52, and of. ii. p. 102, completely corroborates my observations: '• Na Prevlaki neostoje
ni traga rimskomu gradu, a rimski nadpis koji ondje stoji uzidan u crkvici bez dvojbe je iz Bisna iii iz
Kotora doncsen." (There is not a trace of a Roman town at Prevlaka, and the Roman inscription, which
is there walled into the chnrch, has been doubtless transported from Risano or Cattaro.) Dr. Ljubic is
replying to G. Gclchich, who in his Memorie sidle Bocche di Cattaro (Zara. 1880). p. 7, asserts at random
that remains of the city exist at Prevlaka.

6 AiiiiqiKiriaii Eesearchcs in Illyriciim.

Among: tlic coins here brons^ht to lii?lit, I have noticed sovpval silver pieces of
Dyrrluiclnuni and ApoUonia, of the third century B.C., in one case an autonomous
coin of Scodra, datini? probal)ly from about the year 108 B.C.," and I have, myself,
picked up a small brass coin of Bocotia. A few years since there was dug up here
a pale carnelian intaglio in tlic perfect Greek style, representing Apollo Agyieus,
guardian of roads and streets, leaning on a pillar and holding forth his bow.'' "
The old Greek connexion Avitli this ])art of the Dalmatian coast is still traceable
in the local names, and one of the llagu.san islands has ])reserved in a corrupted
form the name of the Elcqihites Nesoi.'^

Finally, I hope to be able to adduce some fresh evidence as to the course of
the land communication between Epitaurum and Narona which may serve to
reconcile completely the statements of Pliny and fbc author of the Tabula Peutin-
f/eriana with the position of Epitaiirum as indicated by existing remains, and may
enable us to dispense once and for all with the ingenious hypothesis of Mommsen.
This evidence I am compelled to reserve for a future paper ; but it may be iisef ul
to mention that I have discovered the traces of the Roman junction road from
Epitaurum, running inland, and not, as hitherto supposed, along the coast ; and
that an inscription on this road shows that, in Claudius's time at any rate, the
maritime terminus of this road was to lie found on the Ragusa-Vecchian site.

The existing architectural remains of Epitaurum are small. The rocky nature
of the soil has hindered tiie usual accumulation of humus, Avliich so often pre-
serves for us at least the foundations of ancient buildings. On the other hand,
what remained of the Roman city has, no doubt, largely contributed to supply its
more renowned mediaeval offspring with building materials. Epitaurum, only
scA'en miles distant, across the bay, by sea, has become a convenient quarry for
Ragusa. Traces of the quay, however, and jiarts of the city walls, may yet be
seen, and the ancient steps, cut in the rock, show that several of the steep and
narrow streets of Ragusa A^ecchia, the small town that now partially occuj)ies the

» Vide Numismatic Chronicle, N.S. vol. xx. pi. XIII. fig. 2.

•> This gem is now in the possession of Mr. W. J. Still man. It greatly resembles that engraved by
King, Aiitique Gems and Rings, pi. XV. fig. 8, and probably preserves the outlines of a celebrated

•= Lopnd {It. Mezzo) in the thirteenth and fourteenth centuries, Z>a/a/o<a, i.e. Da Lafota or D'Alafota,
Cf. Dr. Constantin Jirecek, Die Ilandclstrassen und Bergwerke von Serbien vnd Bosnien u-cihrend des
Mittelallers, Prag, 1879, p. 9. Pliny {II. N. iii. 30, 151), mentions the seven Elapliites Insula; as lying
south of Melita (Meleda).

Antiquarian Researches in Illyricmn. 7

site, follow the Roman street-lines. On the height, now crowned by a chaxiel of
S. Rocco, are evident remains of the Roman cemetery, the oblong cavities of sarco-
phagi being cut out of the solid rock ; and on the shore of the Bay of Tiha, along
which the Roman road leading to the peninsula gate of Epitaunmi must have
run, are still to be seen Roman mortuary inscrijitions cut in the face of a ledge of
roct. That considerable suburbs existed on this side is shown liy the fact that
Roman remains are abundant as far as Obod, where a Hue tessellated pavement '"'
was discovered in the last century ; and in the bay itself walls believed to be
Roman are at times visil)le in the shallows. On the further side of the present
har1)our of Ragusa Vecchia Roman remains are also distinctly traceal)le. In the
walls and courtyards of the present town are fragments of sculpture, and columns,
inscriptions, and monuments, amongst which is an interesting representation of a
Roman Signifer (fig. 1).''

Fii;. 1. Roman Signifer.

•T •■ 1 di lui vivacissimi coluri con maraviglioso artificio fra loio dispusti presentano all' occliio una
serie lumiiiosa di vatrliissiiiu' liste," is Appendini's high-flown description of this mosaic in 1S02. Storia

di Ragusa, p. 50.

•> The on"raving which 1 here reproduce is taken I'rom my work on Bosnia, in wliiili 1 have already
"■iven a popular account of some of the Roman Antiquities of Ragusa Vecchia.

8 Antiquarian Besearches in Illyricum.

Bill llu' most important relic that remains ol lloman Epitaurum is unques-
tional)ly the Aqueduct. The total length of this great work, the remains of which
extend to a mountain soui-ce called Vodovalja, on the further side of the i)lain of
Canali, is about fifteen miles. I have myself traced it throughout the greater
part of its course, and from a comparison of its different levels am persuaded that
the water was in places conducted up eminences a siphon by means of large
reservoirs a ehasse and afiiite, as has been shown to be the case with some of the
great aqueducts of Provence. The arches by which it spanned the level tracts
have unfortunately all perished, though some were existing in the immediate
neighbourhood of Epitaxirum within the memory of man. The last pier of one
of these, formerly existing just outside the present gate of Ragusa Vecchia,
was removed not longer ago than 1875 to widen the road in honour of the
Emperor Francis Joseph's visit. The great length of this aqueduct curiously
illustrates the known daintiness of the llomans in regard to their Avater supply.
At a point several miles nearer Ragusa Vecchia the aqueduct spanned a mountain
source called Gljuta, far more copious than that to which it is ultimately con-
ducted. The water of the Gljuta, so far as my own exi)eriencc goes, is not only
deliciouslv cool to bathe in but eminentlv drinkable. I found hoAvever that the
natives of the district through which the aciiu^duct runs, and to which it gives
its name Canali, the old Serbian Zupa Konavalska, have a prejudice against
either drinking or bathing in the water of this stream. They declare that it
is slightly saline, and that after drinking it you are quickly seized Avith thirst
again, that bathing in it is liable to give you ague, and that it is not beneficial
to herbage. Hence they call it Gljuta, or the bitter Avater. This prejudice may
be traditional, since, although the Canalesi are at the present day a Slav-speak-
ing people, the name Canali itself, and many of the village names" of the district
as well as some of the prevalent physical types attest a coijsiderable surA^Aal of
lUyro-lloman blood.

• As for instance Molunta (cf. Illyrian-Mcssapian suffix -uvtiim, -ventttm, &c.), Vifaljtna from Vitalis,
Cilippi, not to speak of the mediaeval reminiscences of Epitaurum, as Starigrad Pitavr, and its modern
local name, CaWo/^CVi;i<f/<e, cf. Rouman : Cetate, CiV«r, Albanian: Giutet, &c. (cf. p. 32). Excavations
conducted by my friend Dr. I^usdian ami myself in mediseval cemeteries about Mrcinc and 8okk(i, not far
ilistant from the head of the Aqueduct, uiiiply demonstrate the prevalence of non-Slavonic crania. For
the survival of Roman local names in the territory of Ragusa, see Jirccek, op. cit. p. 8. Still more
curious are the fragments of the Roman provincial dialect of Dalmatia existing in the Slavonic dialett
of the Ragusans. Vide Prof. Luko Zore, Dubrovnik, iii. p. 19.^), Nai jezik tijekom vaie knjizevnosti v
Dubrovniku. (Our language in the course of our literature in Ragusa.)

AntiqAia7'ian Researches in lUyricum. 9

The remains of the piers that still exist are formed of a conglomerate of
rul)ble-masonry, mortar, and bricks, and not of deftly-hewn blocks as in the
aqueduct of Salona. The most interesting feature in the existing remains is the
conduit hewn out of the solid rock, which may be traced for miles in the more
hilly part of the country to be traversed, taking great curves in order to maintain
the level. In the last century, to judge from a manuscript letter of the secretary
of the Reimblic of Ragusa, Antonio Alleti, to his friend Mattel at Rome, it must
have been still more perfect. " I have been," he writes on December 14, 1724,
" with much satisfaction at Canali to see the Aqueduct through which the Romans
from a distance of thirty Italian miles [an exaggerated estimate] used to conduct
the water to Epidaurum, and in order the better to enjoy that venerable antiquity
at times I rode on horseback in the very channel in which at one time the water
ran." "

It is noteworthy that in Canali the breadth of the channel of the Aqueduct is
nearly three times as great as at Ragusa Vecchia. More Avater was needed in
this part of its course to be employed in irrigating the fields. The district of
Canali is still the best artificially-watered tract in the whole of Dalmatia, and
the inhabitants seem to have preserved the art of irrigation from ancient days.

The Aqueduct on abutting on the peninsular hill on which Epitaurum stood
ran along the northern wall of the Roman city, which follows for awhile the
northern steep of the peninsula, the city itself lying below on the southern flank
of the hill, where the town of Ragusa Vecchia is at present situate. From the
north-western angle of the old city wall it descends slightly, in part of its course
by a subterranean channel tunnelled out of the rock, to a semicircular Chamber
overlooking the ancient quay, and Avliich appears to have foi-med part of the
public baths.

Just above this spot I excavated a very jierfect portion of the ancient channel.
The channel itself had been hewn, here as elsewhere, in the more rugged part of
its course out of the limestone rock, but the vault above had been constructed of
masonry and concrete. From the pitch of the vaulting to the floor the height

* " Sono stato con sommo contento in Canali per vedere gli avanzi dell' Acquedotto per cui i Eomani
dalla lontananza di trenta miglie avevano condotto I'acqua in Epidauro, e per inaggior godere di quclla
veneranda antichita alia volta con cavallo mi cacciai in quel letto medesimo su cui nn tempo scorreva
I'acqua." The correspondence of Alleti is in the possession of Don Paulovicli of Ragusa, by whose
kindness I am enabled to reproduce the parts bearing on the antiquities of Epitaurum. Cervarius Tubero,
Commentaria suornm temporum, remarks, " Quod autem Canalensis ager territorii Epidaurii fuerit, argu-
mentum est opus mirabilis structurae effectum, qua a vigesimo prope milliario aqua in urbem perducta est,
partim subterraneo rivo, partim opere arquato."



Antiquarian Researches in Ilhjricum.

was exactly five feet, the object beinc^ apparently to enable workmen to walk
along it when repairs were necessary. The rock walls sloped inwards from the
spriniij of the arch so as to present a somewhat coffin-like secti(m, due, no doubt,
as in the case of a coffin, to the desire to give space for the upper and broader
])ai-t of a man's body. The base was trilateral (fig. 2).

Fiji. 2. Section op Aqueduct tunnelled thuouch Rock.

The most remarkable feature, however, is the vaulting above the rock channel.
The concrete with which its surface is coated presents a curious cogged ov
serrated section, due to the impression of the planks of the wooden framework or
centering on the soft material, as is proved by the grain of the wood being itself
in places reproduced. From this it appears that the centering em])loyed by flic

Antiquarian Researches in lllyricum.


Epitaurian architect was different from those generally in use at the present day.
That it consisted of overlapping planks supported below on a semicircular frame-
work is evident, but it is difficult to understand what the special advantages
of this form of centering may have been. The fact, however, that no inter-
stices are left between the planks, shows that the concrete used was of a very
soft nature.

Showing Channel o/ Aqutduct
in -wall and steps of Nldie B

Q, probably Cistern aTid fountain.

E.E. Marble ledge runnings:
round bath

Approx. Breadth of bath floor. U—0=hGft.
Exterior Breadth G_..G.7J/'-


Tlie Aqueduct Channel is Si feet above the cement

floor of bath..
Breadth of outer wall D.F. = 15 feet.

Bath Chamter at Epitaurum.
(Ragusa Veoohia.)

The semicircular basin into which the channel of the aqueduct runs was ex-
cavated by me in 1878 (fig. 3). The water entered the Chamber by a semicircular
niche containing two steps 8 inches high. This again opens into what was
evidently a semicircular Piscina, about 46 feet in diameter, floored with cement,
and surrounded with a ledge on which the bathers could stand. The depth of
the Piscina is 3 feet 6 inches, about half a foot deeper than a similar bath at
Pompeii. Not only the niche and surrounding walls and ledge, but the concrete
floor of the bath itself, had been covered with plaques of marble, all of which —
with the exception of fragments — had been removed by the inhabitants. The
channel of the Aqueduct is continued along the middle of the western wall of the
building, and thence along another wall which follows the line of the straight



Anliquarian JlcsearcJies in Illyricum.

side of the Piscina. Unfortunately, however, the ruin of the rest of the bath
buildini^s has been too complete to admit of reconstruction.

The hitherto known inscriptions discovered on this site ai"e collcct(Kl in the
Corpus Inscriptionitm, and many of those still existini^ on the spot have been
personally examined by Professor Mommsen, The most important of those, con-
taininiif an honorary dedication by the cities of Upper Illyricum to P. Corn.
Dolabella, w lio, as rr()-pra3tor under Tiberius, directed the execution of at least
live great lines of roadway from Salona into the Dalmatian interior, now, un-
fortunately, exists only in a frasj^mentary condition." Accordiui^ to the accounts

Online LibraryArthur EvansAntiquarian researches in Illyricum : Part I-IV Communicated to the Society of antiquaries → online text (page 1 of 32)