Sie sind auf Seite 1von 408

1

=^

"^
())

IJapers

of

tjie

iirc|;raIog}tal

Institute of

%mtxm,

CLASSICAL SERIES.
II.

REPORT ON THE INVESTIGATIONS AT


ASSOS,

1882,

Part

1883,

I.

By JOSEI'II TIIACIIER CLARKE.

Wiii\} an ^ppenliii.

Printed at thk Cost ok the Boston Society of Architects.

NEW YORK:
PUBLISHED BY THE MACMILLAN COMPANY,
66 Fifth Avenue.

1898.

Digitized by the Internet Archive


in

2010 with funding from


University of Toronto

http://www.archive.org/details/p1reportoninvest02clar

hgt

A
aptrs

of

l|t

^rtljirological

|nstittt of Skintrita.

CLASSICAL SERIES.
II.

REPORT QN THE INVESTIGATIONS AT


ASSOS,

1882,

Part
By JOSEPH

aSSit^

1883,

I.

THACHER CLARKE.
an ^ppentiti.

Printed at the Cost of the Boston Society of Architects.

^RCH^OLOGICALN
INSTITUTE

OF
^i^\

AMERICA.
1

879.

/"^^

y/d^

NEW YORK:
PUBLISHED BY THE MAC.MILLAN COMPANY,
66 Fifth

Avenue.

1898.

IX

MICROFILMED BY
UNIVERSITY or 7 ORONTO
LIBRARY
MASTER NEGATIVE NO.:
?3<^/ro

University Press

John Wilson and Son, Cambridge, U.S.A.

ARCH.OLOGICAL INSTITUTE OF AMERICA.


(JTouncil,

1897-98.

PresitJent.

JOHN WILLIAMS WHITE,

Professor

Ph.D., LL.

Harvard

University^ of the Boston Society.

I^anorarg ^rfsltimts.

Professor

CHARLES ELIOT NORTON,

Litt. D., lA^.Y^.^Harifard

University, of the Boston Society.

President SETH
York Society.

LOW,

LL. D., Columbia University,

of

the

New

lJiCE=^resttients.

President

DANIEL

C.

OILMAN,

lA^.Yi.^JoIms Hopkitis University,

President of the Baltimore Society.

Dr.

WILLIAM PEPPER,

LL. D., University of Pennsylvania, Presi-

dent of the Pennsylvania Society.


Mr.
a. RYERSON, LL.B., Chicaoo, of the Chicago Society.
SALISBURY, A.M., LL.B., Worcester, of the
Hon.

martin
STEPHEN

Boston Society.

Professor
the

THOMAS DAY SEYMOUR,

New York

IStiitor

Professor

LL. D., Yale University, of

Society.

in Cl)if of

tf)E

JOHN HENRY WRIGHT,

Journal.

A.M., Harvard University,

of the Boston Society.

tl)er fflcmftors of

Mr.
Mr.

t!)c

(Council.

GEORGE A. ARMOUR, A. M., Chicago, of the Chicago Society.


SELDEN bacon, A.M., LL. B., AVw York, of the Wisconsin
Society.

ARCH^OLOGICAL INSTITUTE.

iv

DAVID L. BARTLETT, Baltimore, of the Baltimore Society.


CHARLES BUNCHER, Detroit, of the Detroit Society.
CLARENCE H. CLARK, Philadelphia, of the Pennsylvania

Mr.
Mr.
Mr.

Society.

Professor

MARTIN

L.

D'OOGE,

Ph. D., LL. D., University of Mich-

igan, President of the Detroit Society.

Professor

HAROLD

Professor

N.

FOWLER,

Ph. D., Western Reserve Uni-

Cleveland Society.

versity, of the

ARTHUR

L.

FROTHINGHAM,

Jr.,

Ph. D., Princeton

University, of the Baltimore Society.

Mr.

MALCOLM

GREENOUGH,

S.

A.

B., Cleveland,

President of the

Cleveland Society.

Professor WILLIAiM

GARDNER HALE,

LL. D.,

University of

Chicago, President of the Chicago Society.

Mr.

CHARLES

L.

HUTCHINSON,

LL. D., Chicago, of the Chicago

Society.

Mr.
Mr.

GARDINER

M.

JAMES LOEB,

York Society.
Mrs. NICHOLAS

LANE,
A.

B.,

A. B., Boston, of the Boston Society.

(Treasurer,) Neiv York,

of the

New

LONGWORTH,

Cincinnati, President of the Cin-

ALLAN MARQUAND,

Ph.D., 'L.U.D., Princeton Uni-

cinnati Society.

Professor

versity, of the

Miss ELLEN
Professor

F.

New York

MASON,

Society.

Boston, of the Boston Society.

EDWARD DELAVAN

PERRY, Ph. D., Columbia UniNew York Society.


de PEYSTER, A. M., LL. B., New York, of the

versity, President of the

Mr.

FREDERIC

J.

New

York Society.
Professor DANIEL QUINN, Ph. D., Catholic University of America,
President of the Washington Society.

Mr. EDWARD ROBINSON, A.


Professor MOSES STEPHEN

B., Bosto?t, of the

SLAUGHTER,

Boston Society.
Ph.D., University

of Wisconsin, President of the Wisconsin Society.


GERALD TISDALL, Ph. D., College of the City of
A'ew York, of the New York Society.
Professor JAMES R. WHEELER, Ph. D., Columbia University, of

Professor FITZ

the New York Society.


Mrs. H. WHITMAN, Boston, President of the Boston Society.
Mr. CLARENCE H. YOUNG, Ph. D., (Secretary,) Columbia University, of the New York Society.

INTRODUCTORY NOTE.

THEbyReport on

the Investigations

made

at

Assos

in 1881,

the expedition sent out by the Archaeological Inissued


stitute in charge of Mr. Joseph Thatcher Clarke, was
The investigations were still in progress, and it was
in 1882.

intended that so soon as possible after their completion a


The work
final Report should be prepared and published.
of the
Most
of
spring
1883.
the
in
ended
expedition
of the
members of the party which had been engaged in it returned

home, and Mr. Clarke at once began the preparation of a


Report designed to give a complete and thorough account of
the unexpectedly important and interesting results of the first
American expedition for archaeological investigation in the
Mr. Clarke's Preliminary Report
field of classical antiquity.
had already given evidence, not only of

his high qualities as

an investigator, but also of his possession of learning ade-

made by the
meet the demands of modern

quate to enable him to set forth the discoveries


expedition

in a

manner

fitted to

scholarship.

During the next two or three years a considerable part of


the work was accomplished and put into type.
The pages which now follow have been ready for publicaBut the publication has been
tion for more than ten years.
delayed, greatly to the disappointment of the Institute, in
hope that the portion remaining to be written might be

completed.

which he was
sonally responsible, Mr. Clarke was compelled

By

a series of calamities, for

in

to

no wise pergive up labor

INTRODUCTORY NOTE.

VI

upon the Report, and

to devote his

whole time to other pur-

From

year to year he has hoped to be able to renew his


labors on the work which it was the object of his just ambisuits.

tion to complete,
It

year

year he has been disappointed.

after

has finally seemed best to the

Council of the Institute

portion of the Report which has been lying

to issue that

ready so long,

in

order that the account which

of the results of the expedition, although

it

it

contains

be but partial

and imperfect, should no longer remain inaccessible. They


have come to this decision with reluctance, both on Mr.
Clarke's account and on account of the Institute.
It is
matter of serious regret that a
the expedition should not be

full

record of the results of

made by

com-

the person most

petent to describe the discoveries and to exhibit their impor-

The Council cannot but hope

tance.

that Mr. Clarke

may

yet find opportunity to conclude his work.


But, unfortunate as the delay in the issue of the Report

has been, the investigations to which

The

interest.

at

peculiar character of

it

relates

many

have not

lost

of the buildings

Assos, and their remarkable preservation, making possible

a complete recovery of the plan

tures quite unique

in

and elevation of

civic struc-

design and plan, give to the work

accomplished there such permanent importance that ten


years

more or

less

in the

date of

its

publication

are

of

comparatively small concern.

Meanwhile

it

is

proposed to publish very

the auspices of the Institute, a work edited

shortly,

under

by Mr. Francis H.

Bacon, the companion of Mr. Clarke in the Assos expedition,


which shall present on a large scale the plans and elevations
of the more important edifices investigated, and shall give all
those details and measurements which may be desired by the
students of ancient art, and especially by those of architecture.
The preparation of this work, in which Mr. Bacon has received the valuable assistance of Mr. Robert Koldewey, who
also took part in the original investigations, will afford material for

full

study of the monuments of various kinds which

Vll

LETTERS.

will be a consite of Assos, and


specially distinguished the

classical

tdbution to

and
archeology of unusual novelty

themselves they
which follow this Note explain
record of the expedition.
form part of the documentary
;

'''T'he letters

C. E.

NORTON.

February, 1898.

Charles Eliot Norton,


President of the

Esq.,

Archmlogual

Institute of

America.

has charged me
Boston Society of Architects
Dear Sir,
of die
President
of conveying to you, as
with the agreeable duty
profesthe
of
the congratulation
Institute of Archaeology,

-The

American
sion

upon

the

work accomplished by your expedition

to Assos

We

Messrs. Clarke and


to thank our brethren,
desire also, through you,
by whose skill,
expedition,
of the
Bacon and the other members
to our
acquisition
valuable
this
of self-denial
energy',

and

fine spirit

knowledge of

rendered
Greek architecture has been

possible.

The

of
Society of Architects to the expenses
contribution of the Boston
concernmg
evidence
new
of
search
Mr Clarke's first expedition in
to know
intended as an expression of its desire
the" Doric order was
the development of Greek archimore of the principles underlying
the complicathese principles, in the midst of
tectural forms

for to

beset all modern works of


and sophistications which inevitably
and rerepair for correction, inspiration,
design, we must continually
logical
the
was
expedition
The later and more fruitful
freshment.

tions

continuation of the

first.

o
. ^,.1,.
grateful to this Society, not only
These successes, therefore, are
o
to justify its first expenditure
because they seem in a manner
spirit
Greek
the
of
exposition
new
means, but principally because this
the most sanguine friends of the

has proved

far

more complete than

us the Greek architect exhad anticipated. It has shown


selfin invention, yet always with
perimenting with forms, and profuse

enterprise

denial

and a

just reserve of force

lesson yet derived from


it

has thrown

new

light

it

has given

Greek antiquity

upon

in

us,

perhaps

the best

the grouping of buildings

art
the divine virtue of simphcity.in

it
;

ARCH^OLOGICAL INSTITUTE.

Vlll

has given us substantially the only examples of the practice of the

Greeks

in

domestic and civic works, and,

in short,

it

seems

brought nearer to our sympathies and comprehension that

modern

the conditions of

We

purifying force.

when

therefore anticipate with interest the

the results of this expedition shall be

and complete

made

its

means

limited

me

edge, has authorized

sum of five hundred

to

to the

policy of contributing to the

its

convey to you

it

its

architectural knowl-

offer to appropriate the

dollars towards defraying the expenses of printing

to raise such additional

pose,

practically available

which we hope may be

advancement of

the Report of the Expedition to Assos,

sum,

if

and

to indicate

may be needed

any, as

being understood that the total amount

thousand

and

moment

as possible, especially in respect to drawings.

This Society, in continuation of


extent of

which

architecture require as a corrective

to the profession in the forthcoming Report,


as full

to have

spirit

is

its

intention

for the pur-

not to exceed one

dollars.

Respectfully yours.

EDWARD
Boston, March

Edward

CABOT,

C.

President.

20, 1884.

Cabot, Esq.,

C.

President of the Boston Society of Architects.

Dear

Sir,

have had the pleasure to-day of laying before the

Executive Committee of the Archaeological Institute of America your


letter to

me

of the 20th instant, and

expression of the gratification which

and of

grateful

their

am

charged by them with the


contents have given them,

its

acknowledgment

the

to

Boston Society of

Architects for the substantial and timely contribution

make toward

the Expedition to Assos.

narrower than

its

the testimony

The income

of the Institute

opportunities for service in

such a contribution
is in

it

proposes to

defraying the expense of printing the second Report ot

is
it

especially

welcome.

Its

affords of the sense of

its field

is

so

much

of work, that

chief value, however,

your Society of the im-

portance of the work accomplished by the expedition to Assos, of the


novel character of the acquisitions

made by

it

in the

domain of Greek

LETTERS.
and of the permanent worth of

architecture,

the

IX

its

results to students of

art.

This testimony, coming from a body so eminently qualified to speak


with authority on the subject as the Society over which you preside,

and expressed by you


it
it

its

has succeeded thus


existence,

upon

tions

terms at once decisive and convincing, while

in

highest degree gratifying to the Institute, as a proof that

in the

is

is

far in

accomplishing one of the chief ends of

equally stimulating to

classical soil

edge of that ancient

undertake fresh investigafurther additions to knowl-

which remains so

art

and

interest to students

to

still

it

which may make

full

of instruction and

the present time.

artists of

The Executive Committee of the

Institute recognize their past debt

to the Boston Society of Architects for

its

contribution to the original

expedition of Messrs. Clarke and Bacon, and their further indirect

but essential obligation to

members of
and

acter

in the fact that

it

the Society,

fact

The Committee

ability.

these gentlemen were

which gave assurance of their charwill

have pleasure in transmitting

a copy of your letter to each of these gentlemen.

They venture

request you to bring the services of Mr. Robert Koldewey, of

who

burg,

some of

has had charge of

to

Ham-

the most important investiga-

hope that it
upon consideration of his part in the joint labors
convey to him a distinct expression of its appreciation

tions at Assos, to the attention of your Society, in the

may

think proper,

on the

site,

to

of the excellence of his work.

The Committee

desire

me

means

to state that the fact that the

for

the publication of the forthcoming Report have been provided by

your Society
will

will

be stated upon

its

titlepage,

be printed as prefatory to the Report

material to be digested,

pared

for the

Report,

and the

will

large

and your

itself.

The

letter to

number of drawings

probably delay

its

me

great mass of
to

appearance for at

be preleast a

year.
I

have the honor to be, with great respect,

faithfully yours,

CHARLES ELIOT NORTON,


President of the Archceological Institute of Atiierica.

Cambridge, March

25, 1884.

CONTENTS.

I.

Page

Course of the Excavations

II.

Acropolis and Temple

40

III.

Temple Sculptures

141

IV.

Date of the Te.mple

Appendix

Relations

292

of

Modern to Ancient

Life

335

AND PLATES.

LIST OF CUTS

Page
1.

2.

3.

4.
5-

6.

7.

8.
9.

10.

11.
12.

13.

14.
15.

Archaic Bronze Arrow-head


Bronze Arrow-head

44
45

Iron Mattock
Present Condition
Plan of the Temple of Assos.
Stone in Foundations of Temple, with Bed-moulds
Isometric
FOR Metal Castings.
Employment of Lifting Dog in laying the lower
Isometric
Steps.
Pry-holds and Levers employed in laying the Steps.
Isometric
Perforation of the lower Step, Eastern Front
Detail of Mosaic Pavement, Southeastern Corner
Outlines of Echinos Curves, Anta Capital, and
Hawk's-bill Moulding of Corona
Upper Surface of an Abacus
Section of the Entablature and Coffered Ceiling

45

OF the Pteroma
Fragments of inner Epistyle Beams, showing Shiftholes and Mason's Marks
Triglyph, Face and Side
Ends of Cornice Blocks, showing Attachments of
B, for
A, for Looped Rope
Derrick Tackle

88

16.

Iron Dog
Cornice Block, as tilted

Turning Grapple
17.

18.

19.

20.
21.

22.

57

58
61

62

67

70
81

83

92

96

in

Lifting.

Release

9^

for
99

Upper
Cornice Block from Southeastern Corner.
Surface and End
Rejected Cornice Block, Recut for Employment in
Tympanon Veil
Beam from the Coffered Ceiling of Pteroma
Beam from the Coffered Ceiling of Vestibule
Beam from the Coffered Ceiling of Pronaos
-.
General Plan of Coffered Ceiling
.

...
.

102

109
115

117
121

123

LIST OF CUTS

xiv

AND

PLATES.
Page

23.

24.
25.

Section of Pteroma
Section of Vestibule and Pronaos
Section of Vestibule Ceiling Beam, showing Lewis

124

Tackle
Fragment of Tile, with Ornamented Edge, from a
Course interposed between lowest Imbrices and
Corona
Antefix. From a Photograph

126

124

26.

27.

28.
29.

3
32
33

130

131
Antefix Section
133
Corners of Imbrices, roughly cut for Jointing
Isometric
134
Constructive System of Pteroma.
From a Photograph
135
Fragment of Gutter.
Section and Scale
135
Fragment of Gutter.
136
Fragment of Ridge Acroterion
Paw of Sphinx or Griffin 137
Fragment of Acroterion.
To face 142
(Plate.) Retreating Centaurs
146
Human-legged Centaur
To face 150
(Plate.) Heracles and Pholos
165
Herakles and the Centaurs of Mount Pholoe
Hind Legs of a Centaur
171
Fragment of a Metope.
(Plate.) Heraldic Sphinxes. Western Facade.
.

31

129

...
...

....

34
35.
36.
37.

38.
39.
40.

To face 172

Heraldic Sphinxes.

Eastern Facade.

41.

(Plate.)

42.

Ionic Capital, with Upright Volutes

To face

434445-

46.

47.
48.

....
....

49.
50.
1.

2.

SI-

73

Couching Sphinx on Sherd from Ophrynion


Figurine from Aqkieui
Squatting Sphinx.
B, Bronze Head of
A, Archaic Coin of Assos.
Griffin, found at Olympia

182

Coin of Assos
Coin of Assos

188

183

187

'89

Mosaic Pavement from a Building South of the


Agora, showing Eagle-headed and Leopard-headed
192

Griffins
48a.

176

Heraldic Sphinx upon

engraved

Seal,

found

at

Assos
Epistyle Relief from the Temple
The Struggle of Herakles with Triton
Epistyle Relief from the Temple
Epistyle Block above the Northernmost Intercolumniation of the Eastern Facade
Epistyle Block above the Southernmost Intercolu.mniation of the Western Facade

200
210
237
241

250
252

Ll^r OF CUTS

AND PLATES.

XV
PaG8

54.

55.

56.

57.

58.

59.

60.

61.

62.

Epistyle Block above the Southernmost Intercolumniation of the eastern facade


Schematic \'iew of the Southeastern Corner of the
Isometric
Entablature.
Epistyle Block above the Central IntercolumniaTioN of the Western Facade
Epistyle Block above the Central Intercolumniation of the eastern facade
Epistyle Block above the Second Intercolumniation
FROM the South of the Eastern Facade
Epistyle Block above the Easternmost Intercolumniation OF the Southern Side
Epistyle Block above the Second Intercolumniation
FROM THE East of the Southern Side
Reconstruction of the Eastern Corner of the Southern Side, showing three Epistyle Blocks relating
TO the Centaurs of Mount Pholoe
Reconstruction of the Southern Half of the Eastern Facade, showing two Epistyle Blocks relating TO THE Centaurs of Mount Pholoe, and the
Central Panel with the Coat of Arms of Assos
Restoration of the Central Acroterion, the Remaining Fragment dotted and shaded
Epistyle Block above the Second Intercolumniation
from the North of the Western Facade
Epistyle Block above the Westernmost Intercolumniation of the Northern Side
Epistyle Block from the Western Group of the
Northern Side
Epistyle Block from the Westernmost Intercolumniation of the Northern Side
Epistyle Block from the Western Group of the
Northern Side
Fragment of an Epistyle Block of the Series relating TO the Erymanthian Boar
Epistyle Block from the Entablature of the Cella,

....

63.

64.

65.

66.

67.

68.

69.

70.

....

above the Pronaos


71.

72.

73.

253

256
261

261

263

265

267

268

269
271

271

273
274
275

276
277
278

Epistyle Block forming the Pendant to that shown


IN Figure 70
Metope of the Eastern Entablature occupying the
Fourth Field from the South
Metope showing the Coat of Ar.ms of Assos

....

279

...

285

285

LIST OF CUTS

xvi

AND PLATES.
Page

74.

"j^.

76.
77.

Metope related in Subject to the Series of the


Erymanthian Boar
Fragment of a Metope, related in Subject to the
Series of the Centaurs of Mount Pholoe
Metope of uncertain Location and Subject
Fragmentary Metope of uncertain Location and

....
....

Subject
78.

80.
81.

82.
83.

of Assos

286

286
286

Plan of the Epistyle of the Temple of Assos, showing IN Black the Position of the known Reliefs
Plan of the Temple of Aigina
Plan of the Theseion at Athens
Plan of the Temple of Assos
Plan of the later Temple of Sounion
Diagrammatic Plan and Dimensions of the Temple
.

79.

285

289
302
303

304
305

320

INVESTIGATIONS AT ASSOS.

CHAPTER

I.

COURSE OF EXCAVATIONS.

IN

October, 1881, the digging of the

was brought

to a close

ruption to which reference

first

year at Assos

by the unwelcome

has been made

official

inter-

in the First Report.*

After the retreat of the would-be commissioner, Mr. Bacon and


Mr. Diller remained upon the

beginning of December.
set

The

in.

By

site

with the writer, until the

that time the winter had

prevalent north winds were so heavy that few

of the small vessels of the country

port of
It

fairly

found their way to the

Behram, while none were ready to venture from

was therefore necessary

to leave the coast of the

Troad

it.

in

the Myzethra, the open sail-boat belonging to the expedition,^

and the passage of the Gulf


with

much

difficulty.

The

of

Adramyttion was attended

little craft,

being heavily laden with

chests of the specimens collected by the indefatigable geologist,

much water over her low gunwale, that two


who had been admitted as passengers, gave up bailing

shipped so

Greeks,

in despair, and,

wrapping themselves

in

their blankets, lay

1 Clarke
(Joseph Thacher), Report on the Investigations at Assos, i88l.
Papers of the Archaeological Institute of America, Classical Series, I., Boston,

1882, p. 44.
2

Report, p. 131.
I

ARCH^OLOGICAL INSTITUTE.

down

some

afforded

and
not

The Musconisi^ and Tokmakia^

in the wet.

Islands

shelter during the latter part of the voyage,

in the harbor

Mytilene the

T)f

full

force of the gale was

felt

This northern wind


Lesbian capital

What he

ble.

is

one of those whose

described by Vitruvius

is

says of

it is,

at least, true in

effect

as

upon the

most deplora-

December

when

the Septentrio blows, the inhabitants do not lounge about the

As

streets because of the biting cold.


built

upon

ran through

neck
its

swampy

of

land,

midst, Mytilene

and as

may

account of Longos

would lead us

in antiquity a canal

and the charming

to believe.

It

possible

is

that

some unfavorable report had reached the Roman

tect,

which

attending

he, in his desire to

coming from the

Still,

interior of this

archi-

exemplify the disadvantages

an unwise orientation

exaggerated and distorted.

is

not always have been so

salubrious as the description of Cicero


^

town

part of the

of streets, has

curiously

the north-northwest wind,

famous and pleasant

and the south wind, wafted across the narrow

strait

island,^

from the

orange groves of neighboring Chios, can hardly have occasioned the coughs and distempers which Vitruvius attributed
to them.

In exceptionally rough weather the regular steamers from

Smyrna

Constantinople do not pass through the channel

to

between Lesbos and the mainland, but put directly out into
the open Aegean from Cape Kara Burnu.^ This being the
ancient Hekatonnesoi, the islands of Apollo Hekatos.
Four small and uninhabited islands lying in the Channel of Mytilene,
the southeast of Cape Argenon, the northeastern point of Lesbos.
1

The

Vitruvius,

Cicero de Lege Agrar.,

'

Longos,

" Insula nobilis et

'

The

I.

I. 6. i.

II. i6.

I.

ancient

amoena."

Cape Melaina.

Tacitus, Ann., VI.

3.

to

INVESTIGATIONS
members

case at the time, the

AT

ASSOS,

1882.

of the expedition took advan-

tage of the passage to Constantinople kindly offered them by


the captain of a Turkish man-of-war, then about to leave the
island.

The

months were spent

three winter

in the preparation of

the First Report, and the drawings which

The work

contained.

it

of archasological investigation

during the second

third years, 1S82 and 1883, was carried on by Mr. Francis


Henry Bacon, Mr. Robert Koldewey, and the writer. Particu-

and
lar

acknowledgment

due to Mr. Koldewey,

an

architect

government and a thoroughly trained archae-

of the Prussian
ologist,

is

who, during

the

first

year of his stay, devoted his

services to the undertaking without remuneration.

The

sur-

veys and restorations made at Assos were, roughly speaking,


so divided that Air. Bacon, besides general topographical work,

investigated the Necropolis, the Gymnasion, and the Greek

Bridge.

Mr. Koldewey was occupied with the Agora and the

buildings in

Greek Bath
chronicle of

its vicinity,
;

including the Stoa, Bouleuterion, and

keeping a general

while the writer, besides

made
Temple and
Theatre and Atrium

the results obtained by the expedition,

all

special studies of the fortifications of the city, the

the

Mosque upon the

Acropolis, and the

of the lower town.

Mr. Joseph

Silas

Diller,

then

Harvard University, returned

holding a scholarship of

to the

Troad

in

1882 for ten

weeks, and completed his geological studies of the country.

John R.

S. Sterrett,

Ph. D., to whose charge the editing of

the inscriptions discovered at Assos had been confided, made,

during

May and

June, 1883, a careful search for epigraphical

materials upon the

stones

previously

Haynes, renewing

site,

while studying also the

removed

to

the port.

inscribed

Mr. John Henry

his voluntary services, took

neaidy

one

ARCH^OLOGICAL INSTITUTE.

4
hundred and

and

fifty

photographs of the antiquities discovered,

of picturesque features of the city

lent themselves

During the

to this

first

manner

and

vicinity

its

of representation.

three weeks of the second year the excava-

tions were under the charge of

Mr. Bacon alone, the

return to Assos having been delayed until the end of

of

writer's

March

Digging was recom-

by the preparation of the First Report.

menced on the 8th

which

March, 1882, with ten men,

a number

gradually increased during the fortnight following to twenty-

They were

five.

set to

work

was

first

in the Street of

Tombs, where the

ornamented sarcophagus (No. XVI.)

substructure of the large

The

freed from the earth.

coffer itself

had been

exposed during the excavations of the preceding year,^ but


the extent and the important character of the

not then become apparent.

been supposed
reality, 2.3

structure

to be the

pavement

metres below

still

The summit

remained

it,

to

monument had

of the pedestal

of the street,

so that

more than one

be excavated.

had

which was,

The

in

half of the
altar

which

adjoined the pedestal was found lying directly upon the pave-

ment, while fragments of the sculptured sides of the sarcopha-

gus were deeply buried

showing that the ancient

in the earth,

Greek road was kept clear from debris


sarcophagus was broken

into.

comparatively late use of the way

at the

time when the

further indication of the

is

the fact that the stones

at the base of the pedestal, before being covered by the earth

washed down from the upper terraces and from the city
walls, had been shattered with a heavy hammer, in order to
extract the lead with which the cramps of the steps had been
set

a vandalism hardly to be ascribed to a time before

arms had come


seems

to

into general use.

The

lid of

fire-

the sarcophagus

have remained balanced upon the broken sides until


1

Report, p. 127,

figs.

33 and

34.

AT

INVESTIGATIONS

ASSOS,

very recently, as the enormous stone,

1S82.

one

of the heaviest

in

Assos, was

Doric column, formerly standing upon the eastern inner

found lying upon the surface of the earth.

corner of the pedestal, lay at a considerable depth, broken in

two by

Several of the steps adjoining the substruc-

its fall.

ture of the tomb, and once leading from the paved street to

the terrace above, were


pithoi (Nos.

6 and

still

Two

in position.

archaic Greek

lying close to the native rock, had, at

7),

the time of the erection of the tomb, been cut through in

digging trenches for the foundation walls of the pedestal and


for those of the terrace

behind

The Greek

it.

builders,

how-

ever, evidently disturbed these archaic jars as little as possible;

the remaining bones, although covered with earth and stones,

not having been moved.


In the

on

first

for little

restricted to

Few

year the excavations in the Necropolis, carried

more than a week, had been almost entirely


the imposing monuments near the main gateway.

sarcophagi were buried in that vicinity after the erection

of the vaulted receiving-tombs,

the

had necessitated the removal of

all

88 1.

The

first

of the

the north of

the large

in the

cemetery dur-

one hundred and twenty-four un-

opened sarcophagi unearthed


at

Hence

earlier remains.

no discoveries of note had been made


ing

foundations of which

in

1882 and 1883, were found

ornamented sarcophagus.

objects in one of these monolithic coffers (No.

three vessels of fine transparent glass, were

2),

among

The

notably

the most

made by the expedition.


March a number of trenches were

valuable discoveries of the kind

Towards the end

of

opened on the lowest


lis,

the western side

and were subsequently carried across

a northeasterly direction.

and the position


determined.

The

of the Necropo-

all

the terraces in

original levels of the street,

of the larger burial enclosures,

were thus

many

ostothekai,

This digging brought to light

ARCHAEOLOGICAL INSTITUTE.

These

containing the crumbling fragments of burnt bones.

remarkable urns, of great age, were found only within a limited area,

and always rested

directly

upon the native

rock.

So closely together did they lie, and so delicate and fragile


was the pottery of which they were made, that it was necessary to use knives for the removal of the earth around them,
two specimens

after the

first

found had been shattered by the

heavy picks.

mean

In the

while, plaster casts of those temple sculptures

which had been discovered during the first year were made
by a marble worker from the island of Tinos, Jani Laludis,

who had been with Dr. Humann


of these casts

Three

Pergamon.^

were prepared, and forwarded respectively

Museum

Boston

at

of Fine Arts, the Louvre, and the

sets

to the

Museum

of

But as the only plaster to be procured was


and the inadequate appliances at hand could not prevent the warping of the glue moulds, the results were far
from satisfactory. This is the more to be regretted, as it is
of inferior

Berlin.

quality,

now

obtain

impossible to

new moulds from

rehefs

those

which, in the official division, fell to the share of the

Turkish

government.
Several days in

suspended.

On

hail drove the

March were

so cold that

the 15th of the

workmen from

month

work had to be
snow and

a storm of

the trenches, and even as late

as the 8th of April the temperature was so low that standing

water

in

the lowlands of the Troad was covered with

The want

of comfortable quarters at

Behram during

ice.

this

inclement season, together with the attractions of the Easter


festival

upon the island

of Mytilene, caused the

Greek

labor-

and excavations could not be


Durtheir return on the 17th of April.

ers to desert the site in a body,

recommenced
1

Humann

until

(Carl),

Die

Ers^ebnisse der

der Untcrnehmung, Berlin, iSSo, p. 20.

Ausgrabungen zu Pergamon.

Geschichte

INVESTIGATIONS

AT

ASSOS,

1882.

ing this interval the surveys and measurements were


gently carried on.
foot

by

foot,

The

entire field of ruins

dili-

was searched

and the writer had the pleasure of finding the

third block of the sphinxes from

the western front of the

temple, lying half buried in the earth, face downwards, upon

New

the slope of the Acropolis.

by carpenters

in Molivo,

wheelbarrows were made

and the blunted pickaxes were sharp-

ened by a gypsy blacksmith who had encamped in the neighboring village of Pasha-Kieui.

After the Easter holidays so


it

to

was often found impossible

make

many men were engaged that


enough small money

to collect

out the weekly wages.

of this primitive country

is

great part of the business

carried on

by

barter,

and

the

all

small coins of silver and copper obtainable from the bakhals


of

Behram and

the neighboring villages were not sufficient

for the needs of the expedition,


to disburse from one
jids in small

sums.

obliged

on every pay-day

hundred to one hundred and

As

fifty

the expedient of paying several

med-

men

together with a gold piece proved unsatisfactory, the example


of the

parochial churches of Mytilene was followed, and a

quantity of paper money, of small denominations, was issued

by the expedition.

The

bits of

green cardboard, signed and

stamped, were readily accepted, and circulated so widely

throughout the southern Troad that some

difficulty

was ex-

perienced, at the close of the work, in calling in the out-

standing amount.

When

the digging was recommenced, the entire force was

employed upon the terrace before the Stoa.


the western end of the

The temple

at

Agora was thoroughly examined, and

the position of the neighboring streets and pavements deter-

mined

suflficiently to

enable Mr. Koldewey to begin his detailed

survey of the Stoa and the adjoining buildings.


ble pedestal of a statue with an inscription to the

The marEmperor

ARCHAEOLOGICAL INSTITUTE.

337-340) was found, during these investigations, lying buried beneath the debris accumulated in
Constantine

(II.

a. d.

the street upon the north of the temple.

made

new road was

from the eastern side of the Acropolis to the port, and

the sphinx relief was dragged

Awaiting the advance

removed

the

to

enclosure was

down upon

the sledge.

men were

of the survey, the

Necropolis,

where the

thoroughly excavated.

sulted in the discovery of

again

so-called

Larichos

The work

here re-

numerous sarcophagi and cinerary

urns, the former containing pottery, glass, strigils, coins, and

some few ornaments

of gold

and

silver.

Together with these

were the two best preserved and most ancient skulls found
Assos,

discoveries

of far greater value to science than could

The

have been the richest treasures of precious metal.


antedating the Persian war, was

one,

in a large pithos (No. 5)

the other, referable to the second century


lithic

at

b.

sarcophagus (No. 32) of the Larichos enclosure.

eral inscribed stones

were also unearthed.

On

April, thirteen, and

on the 24th, no

than

less

c, in a mono-

Sev-

the 22d of

seventeen

previously unopened sarcophagi were brought to light.

In the following week excavations were resumed upon the


Acropolis, where was found the largest of the epistyle reliefs
of the temple,

the four centaurs with horses'

workmen remained upon

fore-legs.

The

the Acropolis until the 13th of May,

the digging being further rewarded by a second block of the

centaur

relief,

the

paw

of

the acroterion griffin, an important

fragment of the ornamented terra-cotta gutter of the temple,

and an inscription containing an inventory of the chattels


the building

Greek

of

itself.

festivals occurring

during the

rupted the work for several days.

first

part of

May

inter-

This opportunity was

taken by Mr. Koldewey and by the writer to

make

a journey

through a previously unvisited tract in the interior of the

INVESTIGATIONS

AT ASSOS,

1882.

Troad, lying between Assos and Lecton on the south, and

Alexandreia Troas and Neandreia on the north.

towns

ruins of the three

last

Besides the

mentioned, those of Sminthe,

Tragasa, and Larissa, were explored, as well as those of sev-

names

eral ancient vilUages the

especial attention being devoted

of the site of Polymedion,

locality,

later

their bearing

to

The most important

remains at Assos.

A month

which are not known,

of

upon the

discovery was that

on the coast, opposite Methymna.

instruments were brought to this interesting

which was carefully surveyed, some digging being

necessary in order to follow the circuit of the fortifications,

and

to ascertain the extent of the sacred

grove which occu-

pied the summit of the Acropolis in place of the customary


temple.

Eight days were devoted

Polymedion, the results of which

to these investigations at

be given

will

in

a separate

publication of the Archaeological Institute.

On

the 17th of

May

excavations were begun at the east-

ern end of the Agora, the foundations of the Bouleuterion

being

laid bare, while the stairways

were cleared.

mass

debris

of

Greek cistern on a lower

at

the south and west

was removed from the

terrace, the existence of

become known during the

first year.^

ranean vault were discovered,

Within

which had

this subter-

in a fine state of preservation,

the marble head of a heroic statue, and several additional

fragments of the inscribed


First Report.^

with

many

The accumulated

through the narrow

and was carefully


1

Report, p. 37.

Report, Appendix,

of Assos

published as No. 3

in the

earth was found to be mixed

sherds of water vessels of the Byzantine period,

and with the bones

tions

stele

of

domestic animals.

orifice

by means

sifted in the

No.

3.

open

It

was removed

of baskets

and ropes,

air.

Also, Sterrett (John Robert Sitlington), Inscrip-

Archaeological Institute of America, Boston, 1S85, No.

XXVIII.

ARCHAEOLOGICAL INSTITUTE,

lO

The

Bouleuterion proved to be a construction of

terest and importance.

eral inscribed stones,

Upon

some

of

its

much

in-

plan were discovered sev-

which had been

diao-onal walls of a late restoration.

built into the

All of the

men

could

not be employed upon this spot, and a part of the gang was
removed to the front of the Stoa and to the terraces below its
retaining walls.

The

detailed examination of the long col-

onnade, the place of assemblage before the bema, and the


ramps and steps leading to the upper town, was a work of
great extent, which thenceforth received uninterrupted attention for

more than a

year,

two or three men being always

here employed to clear the pavements, stairways, pedestals,


and water-courses, and to aid in the surveys and measure-

ments.

The

intimate acquaintance thus obtained with the

closely connected group of structures surrounding the market-

place has proved to be one of the most important results of

the investigations.

During the

latter

days of

May and

the

first

week

of

June

the greater part of the force was engaged at the theatre, the

thorough investigation of which occupied twenty men for


The marble columns which supported the
three weeks.
stage, the water-works for cooling

and draining the enclosure,

and both the vomitoria, were thus discovered, while a considerable extent of the seats and passages of the auditorium,

and of the encircling stairs and streets outside the structure,


were freed from earth. On the completion of this task most
of the laborers were again set to work in the Necropolis.

Only a few remained within the town, where, on the loth of


June, they had the good fortune to bring to light the longest
inscription found at Assos, buried beneath the pavement of
a Christian apse built into the small temple at the western

end of the Agora,

While excavating those m9numental tombs and

burial en-

INVESTIGATIONS AT ASSOS,
closures which were situated at

1SS2.

some distance from the

prin-

unopened sarcojDhagi (Nos. 52 to 82) were


The number of cinerary urns was by this time nine-

cipal gate, thirty

found.

teen

of

The

seven.

pithoi,

quantity of small articles

and the neckband,


series, all

ring,

coffers

contained the usual

the figurini, the coins of Assos,

and beads

of gold, belonging to this

being of especial value.

Notwithstanding the heat and

stifling dust

of July and

August, the excavations were actively carried on, the number


of

men being

gradually increased to forty-one.

more.

The

outfit

and wheelbarrows

of the expedition did not furnish picks

for

After the 27th of June this force was directed to the

most extensive task

of the undertaking, namely, the

thorough

examination of the enormous mass of earth and stones which

had accumulated beneath the terrace of the Agora, between


its

For

retaining wall and the upper seats of the theatre.

nearly ten weeks, until the 9th of September, the whole attention of the expedition

was directed

which much had been expected.


probability that

many

enormous heap

of rubbish,

works of

art

from

antique remains would be found in this

where

all

the public records and

which must once have stood upon the Agora and

in the adjoining

by

to this locality,

There was, indeed, every

buildings would

pillagers of the city.

naturally have been cast

The experience

vestigators upon ancient sites had

shown

of all previous in-

that considerable

deposits of antiques, especially fragments of sculpture


inscriptions, existed in the chutes

and

formed by the overthrow

monuments adorning such centres of civic life.


the earth beneath the Agora of Assos had been

of the smaller

In so far as

examined during the

first

year, the results

had borne out

this

Almost all the inscriptions published in the


first Report, among them the valuable bronze tablet with the
oath taken by the Assians on the accession of Caliglila, were

presumption.

ARCH^OLOGICAL INSTITUTE.

12

found in the ruins of Byzantine buildings, situated just below


But, however well grounded, these expec-

the Bouleuterion.
tations

were almost entirely disappointed, during the long

work of the second

The movable

year.

marble carvings, inscriptions, and the

objects discovered,

like,

were few,

in

themselves not sufficient to repay the expense of the work.


In architectural respects, on the other hand, this

eminently productive.

At

paved with exceptionally

was

field

the east were found two rooms,

and early mosaics, the one rep-

fine

resenting the coat of arms of the

city,

two crouching

griffins,

the other a vender of Cupids, with Nikes and tripods at either


side.

The monumental

below

to the

middle

flight of steps leading

of the

from the street

Agora, the Heroon, and, above

the unique Greek Bath, with

its

all,

three stories surmounted by

a broad colonnade, must be considered

among

the most strik-

ing results of the expedition.

Five of the nine weeks were devoted

to this edifice.

The

greater part of the cisterns, and three of the lower chambers

This was a work of considerable

were thoroughly excavated.


difficulty, as

the heavy stones of the superstructure entirely

Near the northwestern

covered the plan.

corner, on the level

of the Agora, was found the remarkable standard of roofing

but no objects of interest were brought to light within

tiles,

the building

On
man
by

itself.

the south of the adjoining street the remains of a

bath were discovered.

their continued

pecially

These had been

employment during Byzantine

by the reconstruction of the chief

church, to warrant the expenditure of


in

their exploration.

which appeared
the

rest.

tions,

to

too

Still,

Ro-

much

injured

ages,

and es-

halls as a Christian

much time and

labor

four chambers were excavated

have been buried

at

an

earlier

epoch than

In them were found the fragments of two inscrip-

dedicatmg the bath and

its

belongings to Julia Aphro-

INVESTIGATIOXS

AT

ASSOS,

1S82.

and thus giving an accurate date for the construction.


The workmen highly approved of researches in this vicinity,
large coarse
as they were allowed to carry off such of the
paving tiles and drain pipes as were of no value in the invesAt the end of the day each of the Greeks went
elite,

tigation.

down

much of this earthenware


The tiles were used
shoulders.

to the port laden with as

he could carry upon

his

pave bakers' ovens

the pipes, as gutters for roofs.

as
to

When

island of
the writer passed along the northern coast of the
to
Mytilene, some months afterwards, every village seemed

be thus provided.
susIn the caldarium of the bath the space between the
filled with
pensura?, beneath the floor, was found to be still
by the
air
the
into
fine wood ashes, which, being whirled

high winds, covered everything in the neighborhood with a


It was a picturesque sight, at nightfall
thick white coat.
work in this locality, to see the men standing in a long
after

row on the

large stones of the ancient

mole which

still

pro-

their evening

Here they washed before

above the water.


meal, which, like the laborers of classic antiquity, they not

ject

unfrequently ate in the dark.


the 4th of September, as the funds at the disposal of
the expedition ran low, it was necessary to dismiss a great
week later, those remaining were
number of the men.

On

Gymnasion and

transferred to the
1

8th,

many

of these

had

also to

be sent away.

ber only seven

men were employed,

banks of

and

earth,

urements.

its vicinity

chiefly in

in aiding the surveys

but on the

During Octoremoving small

and detailed meas-

In an undertaking directed merely towards treas-

ure-trove, such a diminution in the

have been equivalent

was not the case

to

of laborers would

an entire cessation of work.

at Assos.

for the explorers, released

number

The

delay rendered

from the superintendence

it

This

possible

of the dig-

ARCH^OLOGICAL INSTITUTE.

14

ging, to carry on the investigations necessary for determining

The members
more busily or more prof-

the character of the most recent discoveries.


of the expedition were at no time

employed than during these weeks, when the


previously obtained were collected and systematized.
itably

Assos

in this respect presented peculiar difficulties.

degree of

demolition was such as

sight, that architectural

hopeless.

results

The

to

make

seem, at

it

The
first

investigations were here altogether

walls within

the

city

had everywhere been

levelled to the present surface of the earth, and in those cases

where the buildings were elevated upon

artificial

terraces the

foundations themselves had been washed away by torrents of

Throughout the entire city, less than half a dozen


columns were still erect, and even these were without entablaNot one stone remained in position above
tures and capitals.
Retaining walls and ramparts,
the steps of the great temple.
sufficiently heavy to withstand the wanton destruction of man,
winter rain.

had been thrown out of position by the many severe earthAn enormous mass of
quakes which Assos has experienced.
masonry, for instance, bordering the Agora upon the south,

overhung the bath by more than

half a

metre

while the bed-

joints of a fortification wall three metres thick,

of the eastern enclosure, were

lifted to

forming part

an angle of not

less

than fifteen degrees.

The

vestiges which had survived this terrible demolition

were buried beneath stones

fallen

from the upper part of the

buildings, and generally also beneath

some accumulation

of

This had been overgrown by dwarf oak bushes, intertwined with briers, and as these are the only forms of vegetation spared by the browsing goats and camels, they had

earth.

covered the heaps of debris with low, impenetrable thickets.


Such was the aspect of the entire site on the writer's first
visit to

Assos

in

1879.

INVESTIGATIONS
The work

of recovery

AT ASSOS,

was begun by burning the bushes.

In the dry season the tangled mass took

fire readily,

roaring and crackling flame quickly swept


of green

15

1882.

which covered the heaps

of

away

moss-grown

and the

the patches

The

ruins.

formless rubbish was then removed, and, when the position


of the ancient walls became recognizable, trenches were dug

on either
block

side to determine their character

retaining

still

its

original shape,

not to the edifice upon whose plan

it

and extent.

was found, was meas-

ured and drawn to a uniform scale, generally

wonderful how

this

"

Every

whether belonging or

20.

order gave each thing view."

It

was

While

the plan of a building could be followed by the foundation


walls,

if

not by marks upon the pavement, the elevations

were recomposed upon paper,


brought to

The

light.

bit

by

bit,

from the fragments

height of the columns, and conse-

quently also that of the stories in which they were employed,

became evident from a comparison of the proportional diminution of all the drums with the diameter traced upon the
stylobate,

holes

for

and with that

of the

necking

of the capitals.

The

dowels and cramps of metal provided the most

absolute proof of contiguity

separate stones in courses

and even the position of the

long overthrown could be deter-

mined from the shift holes which it was customary throughIn


out Greek antiquity to cut upon the beds beneath them.
short,

it is

not too

much

to say that

one intimately acquainted

with the architectural methods and details of the ancients

can reconstruct their edifices with absolute certainty through


a close study of overthrown and widely scattered stones,
just as a naturalist,

from a handful of

fossil

bones, can pre-

sent the image* and describe the very habits

which

for

of

an animal

thousands of years has had no living represent-

ative.

The

task of tracing the connection between the architec-

ARCH^OLOGICAL INSTITUTE.

tural

members was complicated,

in

this case,

having been scattered over the entire

and Turkish

builders.

by

site

through their

later

Byzantine

Stones of the upper story of the Stoa

had been used for the Christian church on the terrace below
the Agora; a capital, an entablature, and the lintel and jambs
of a door

from the lower town, together with

many

blocks

from the summit of the Acropolis, had been built into the

mosque

beams

ceiling of the great temple

of the coffered

had been employed

in late structures that

stood at the east of

The

the Bouleuterion, and at the south of the Gymnasion.

together of such disjecta

fitting

membra

in

some cases

in-

more than a thousand measurements. The homogeneous character of the material was the source of even
greater difficulty.
Without a single exception, the buildings
of Assos, from the archaic Greek temple to the most recent
volved

hovels of

Behram

Thus, while

in

village,

were

second andesite.

built of the

the investigations

instance, those of the neighboring

among

other ruins

Pergamon

grain of the various limestones were

among

To

this

may be added

andesite, although in general suffering but


ing, is easily

chipped and

were frequently broken

split,

for

the color and


the

most readily

recognizable and trustworthy indications, at Assos


indistinguishable.

all

was

the fact that the

little

from weather-

so that projecting mouldings

off altogether.

During the second and

third

weeks

of

September, the

writer profited by the presence of Mr. Diller to visit with

him a large part

Theban

plain

between

this

route included
gulf.

to

of the western

and Kisthene
all

known

as

Aphrodisias.

The

the coasts of the mainland bordering the

rapid survey was

referred to in the

and southern Troad, the

the south of Adramyttion, and the tract

first

made

of the ruins

on Qozlou-dagh,

Report as Lamponeia, and a remark-

able fastness was discovered

upon the very summit of Mount

AT

INVESTIGATIONS
which

Ida, enclosing a spring

ASSOS,

1882.

but a few steps from the

rises

These explorations, interesting

highest peaks.

\*j

alike in geo-

and topographical respects, were continued beyond the


head-waters of the Aisepos, to an ancient site where various
logical

fragments

of

marble sculptures

in relief

were found, and sub-

sequently removed to Assos.

On

the 26th and 27th of September, while the digging was

being carried on with but few workmen, a number of most


'

welcome guests were entertained at Assos


Prof. W. W.
Goodwin, of Harvard University, then on his way to Athens
as first director of the American School, Prof. R. C. Jebb, of
:

the University of Glasgow, Frank Calvert, Esq., and three


ladies.

Professor Jebb has published an interesting account

of his journey

also

through the Troad on

some comment upon the work

This opportunity

may

this occasion,

at

making

Assos.

pass without mention of the

not

obligations under which the expedition stands to Mr. Calvert.

His friendly assistance was given on

undertaking

to the

the exploring

party

well-known farm

at

who was

possible occasions,

member

so fortunate as to visit

Thymbra, or the hospitable house

A familiarity

Dardanelles.

all

well as to every

itself, as

with

all

of

the

at the

parts of the Troad, com-

bined with exceptional interest and information in various

branches of
est value.

scientific research,

To

say this

is

rendered his aid of the great-

but to repeat the testimony of every

scholar of our generation

who has worked

in

this

part of

Asia Minor.2
1

No.

Jebb (Richard Claverhouse),


London, 1883.

Tour in

the

Troad.

Fortnightly Review,

CXCVL,

2 Stark (Carl Bernhard), Jenaer Literatnrzeitung, Jena,


1877, No. XLIV.
"Jeder derseit Jahrzehnten
Gelegenheit gehabt hat an den Dardanellen und
in der troischen Ebene zu weilen, kennt den Namen der Familie Calvert, und
weiss dankbar zu riihrnen was besonders Frank Calvert durch immer neue
Untersuchungen und durch uneigenniitzige Unterstutzung und Berathungen
der Reisenden der Erforschung jener Gegend und ihrer Alterthiimer geniitzt
.

ARCH^OLOGICAL INSTITUTE.

Assos

is

more frequented highways


The next were two very humble

so far aside from the

that visitors were rare.

German Handwerksburschen, brewers by

trade,

who were

pur-

suing a somewhat devious course homeward from Smyrna.


In default of the wonted ZeJirpfennige of their native villages,

they had supported themselves on the road by peddling small

packages of polishing-powder of their own manufacture.

had walked

were glad

all

to

the way from Persia, across Asia Minor.

work

for a

One
They

time in the survey, and made them-

selves exceedingly useful.

After the 6th of November, funds having been received,

was again engaged, and the digging


Gymnasion was at last completed. Towards the close

the

full

force

month the men were divided


ployed upon the Agora,

into small gangs,

in the

to

end

all

the

of the

and were em-

Tombs, and upon


was then thought it

Street of

As

several parts of the fortifications.

might be necessary

at

it

excavations at Assos with the

season of 1882, every exertion was made to complete the most

important investigations before the advent of that midwinter

month of Lenaion, the cold of which is as terrible to the


modern as to the ancient laborer.^ The men dug on Sundays, feast-days, and even on

Christmas,

superstition of the Greeks certainly


tion

and their love

of

money.

It

is, it

for,

great as the

yields to their ambi-

was not even found neces-

sary to increase the wages on these days, as had been done

hat."

Compare

also

the

same writer

in

Nach dem

his

griechischen

Orient,

Heidelberg, 1874.
Prof. Dr. Ascherson, director of the Botanical Museum of Prussia, says, in
his Beitrag zur Flora des nordwestlic/ten Kleinasiens {Jahresbucher des Botanihheii Aluseufns, Berlin, 1883):

"Calvert, dessen vielseitiger wissenschaftlicher

Bildung und lebhaftem Interesse Naturwissenschaften und Archaologie schon


."
manchen dankenswerthen Beitrag verdanken.
.

'

See Hesiod's

Days, 504-563-

fine description of the rigors of

Lenaion

in the

Works and

INVESTIGATIONS AT ASSOS,
by Dr. Schliemann

at the

1882.

neighboring Hissarlik

tians at

Assos were unable

as the

Mohammedans

demand such

to

had, from the

first,

19
^

the Chris-

a discrimination,

worked

for

the

usual pay on Fridays.

The orthodox Greeks


working on the many
declaring that,

if

had, indeed, far greater objections to

days than on Sundays, always

saints'

they should

fail

question would do them some

in

this respect

were very decided.

in

honoring him, the saint


Their convictions

evil.

It

in

happened that on such a

the year, the writer was helping a numworkmen to move one of the inscribed epistyle
blocks of the Bath, when it fell upon his foot, which was so
crushed as to prevent his walking for three weeks.
The
7rav}]yupi<;, earlier in

ber of Turkish

accident was
that
for

regarded by the Greeks as a clear evidence

the offended

ill.

It

saint

proved of but

had interceded, not for good, but


little

moment,

a roundabout ascent, the field of ruins

since,

by following

could be visited on

horseback.

That the greatest care was taken


intending the work
spite of the

danger

will

of

in laying out

and super-

be evident from the fact

that,

in

digging in deep pits and trenches, from

the sides of which enormous beams of stone often projected, no

One Greek

serious accident occurred during the three years.

laborer

was knocked down by a

slide of earth

beneath the

retaining wall of the Agora, and, as his complaints were so


pitiful that internal injuries

were feared, he was

at

once taken

across the strait to the village doctor of Skamnia.

returned in a fortnight, asking to-be employed again.

But he

Even

apart from the relatively greater risk of the excavations, this

compares favorably with the general


and railroad building.
^

The

official

statistics of

earthwork

records of France, for in-

Schliemann (Heinrich), //ios, City atid Country of the Trojans, London,


and Troja, London, 1S84, p. 11.

1880, p. 661

ARCH^OLOGICAL INSTITUTE.

20
Stance,

show an average

of

one individual maimed

in

the

expenditure of each ^40,000.

The Greeks were


hardy, in

the work

and even somewhat

light-spirited,

fool-

but the Turks, while of greater strength

and bravery, were more quiet and

careful.

The

latter

were

always chosen for posts requiring especial steadiness and endurance, such as the tottering upper courses of the towers

and

walls,

and the narrow

sunk between overthrown

pits

blocks in order to examine the

sills

of the great gateways.

After the trustworthiness of Omer, the head workman of the

Turks, had become known, he was permitted to carry the

heavy and extremely


to place, over

delicate transit instrument from place

heaps of rocks and up steep ascents.

His

strength and fineness of touch in this responsible task were

remarkable.

Notwithstanding

all

made

the efforts

after the arrival of funds,

it

was

still

to hasten the

work

found impossible to bring

the investigations to an entirely satisfactory conclusion by the

end of the second season. The removal of the deep earth accumulated beneath the Agora had occupied the busiest months
of the year, and had required

more time than could have been

foreseen in laying out the work.


small force of

had been able

men employed
to

As

has been explained, the

during September and October

accomplish but

And

little.

yet

it

was the

intention of the promoters of the undertaking, as well as the

great desire of those intrusted with

nothing henceforth

to

its

be done upon the

execution, to leave

site of

Assos,

even

by the most careful gleaner.


Therefore, in a letter addressed to the committee of the

Archaeological Institute shortly before the suspension of the

excavations in

December,

1882,

it

was recommended that

the work should be prosecuted during a third season,

as

long as was permitted by the trade, which had been granted

INVESTIGATIONS
May, 1881,

in

for a

of this proposal

were then,

of the

21

18S2.

The

responsibility

threw a great burden

It

felt.

committee, who, with limited means,

addition to the

in

ASSOS,

term of two years.

was seriously

upon the members

AT

work

at Assos,

carrying on

American archaeology.
The considerations upon which the recommendation was

extensive researches in the field of

Much remained

founded were, however, of decisive weight.


be freed from

to

investigations could be

earth before the

regarded as absolutely thorough, and the recovery of the


ancient city as complete as

Agora

it

was necessary

Upon

made.

could be

it

between the Stoa and the Bouleuterion, south

lated

the

remove a mass of debris accumu-

to

the

of

great flight of stairs, in order to determine the character of

monument

the
to

which

and the juncture between

be examined.

subsequently proved

The ends

be the chief bema of the town.

also,

at

there existing,

of the reservoir,

and the Stoa, were yet

it

Assos rendered the thorough investigation

its

to

The unique importance of the market-place

vicinity a matter of the greatest

of all points in

moment.

As

it

had

al-

ready proved to be the most complete and interesting Greek

Agora known, no stone should be


throw further
the buildings

light

surrounding

extent was the work


the

city.

No

left

still

to

it.

Equal

in

importance and

be done upon the fortifications of

digging had hitherto been attempted at several

of the gates of the ancient enclosure.

used by the

unturned which could

upon the arrangement and appearance of

They were constantly

Turkish inhabitants of Behram, and

it

was

thought advisable to defer the trouble which must arise from

any interference with these thoroughfares


close of the undertaking.
sufficient to

weeks
city,

the

At

until

towards the

the Gymnasion, work remained

occupy a large body of men for two or three


same was the case with the main street of the

between the great eastern gate and the Agora.

The

ARCHAEOLOGICAL INSTITUTE.

22

exceptionally well preserved ruins of a structure in the lower

map

town, called, upon the

had not been examined at


at the Necropolis

quently proved to be

To

of its kind.

this

for publication

first

And,

year, a

finally,

were incomplete, there

some distance from the

fit

of the

all.

city walls, a

among
it

Roman

portico,

the investigations

remaining, at

still

mausoleum, which subse-

the most interesting structures

must be added,

that no photographs

had been taken during the past summer, as

the gelatine was frilled by the great heat

and

also, that

no

professional epigraphist had visited the site to examine those


inscriptions

which could not be carried away

at the close of

the work.

Circumstances of recent occurrence, not immediately connected with the undertaking, gave exceptional emphasis to this

recommendation.
had, a

short

The Turkish Ministry

of Public Instruction

time before, annulled the existing laws con-

cerning excavations within the limits of the Ottoman Empire,

and had resolved thenceforth


to excavate,
all

and even

to grant

no further permission

the sale and exportation of

to forbid

antiques discovered in the Turkish dominions.

tute

upon

was engaged
classic

mains could
restricted

soil,

in its last, as well as its first

thus

It

appeared more than probable that the Archaeological

Insti-

undertaking

from which the acquisition of ancient re-

be hoped

freedom of

the

of Greece

laws

archaeological

bidden the export of antiquities.

having long

investigation,

This made

it

and

for-

especially

desirable that the explorations at Assos should be completed

with the utmost thoroughness,

even

though the resources

up for some years to come.


by the committee, and
was
adopted
The recommendation
the requisite funds were promptly subscribed by a number of
gentlemen interested in the progress of the work. It was

of the Institute should be taken

soon learned by telegraph from Boston to Assos that ample

AT

INVESTIGATIOiXS
means had been provided

for the

ASSOS,

1883.

23

continuation of the excava-

tions during a third season.

During the

first

three weeks and a half of December, the

men, gradually decreasing

staff of

in

number from twenty-

three to twelve, was employed in the Street of Tombs.


the progress was not

rapid, as

many

with the work, and water stood deep

On

in

But

rainy days interfered

the pits and trenches.

month Mr. Koldewey and the writer left


Athens. Digging and sledging went on for ten

the 26th of the

the site for

days longer, under the superintendence of Mr. Bacon, after

which time the investigations were suspended.

weeks

before, the streets of the little village

at its port

had ceased

to

six

and the landing

be a place of assemblage for the

country people.

The

smoky

huddled together over basins

interiors,

Even

patrons of the various cafes sat in the

burning

of

The doors of the windowless hovels, always open


during the warm season, were now tightly closed
within
hibernated the women and children, wrapped in the gaily
charcoal.

colored rugs which they had

woven during the long

rain-

The very dogs had hidden themselves away, seeking shelter in corners of the many unoccupied
houses which attest the greater extent of Behram in former
To one riding into the squalid village during this bitages.
storms of the early winter.

terly coJd season, the place

ment

seemed uninhabited,

the

settle-

of the Turkish conquerors itself a ruin.

The

writer returned to the Troad alone, on the 28th of

The beginning of the digging was delayed


more than a week by the slowness of the Kaimaqam of

January, 1883.
for

Alvadjyq

appointing a successor to the

in

official

supervisor

who was prevented by illness from resuming his


The post was ultinfktely assigned to Hadji Chris-

of the work,
functions.
tos,

the Greek merchant living at the port, whose friendly

24

ARCH^OLOGICAL INSTITUTE.

service to the

members

at the site, has

With eighteen men,


this

that could be brought together at

all

town designated upon the plan

Roman

port as a

As

portico.

of the first

Re-

the excavations advanced,

proved to be the atrium of a large palace-like dwelling.

With exception
mosque,
but

first arrival

to.^

time of the year, work was commenced at that structure

of the lower

this

of the expedition, on their

been referred

it

it

of the Christian

was the building

churches and the Turkish

of latest date

was well constructed, and,

examined

in design,

at

Assos

proved of interest

as exemplifying the persistent retention of Hellenic forms


late into the ages of

example

Roman

furnished an additional

It

rule.

of the civic architecture of the Greeks, the devel-

opment and adaptability of which is so well shown by the


monuments of Assos.
During February the work was carried forward under great
difficulties.
The loth of the month was the coldest day of
the year, and a fortnight

later

there were

long-continued

storms of hail and snow, which put a stop to


Nevertheless,
of the

it

was found possible

Atrium so rapidly as

all

digging.

advance the excavation

to

to allow a part of the

gang

to

be

transferred to the Stoa, and to the small aediculas at the west


of that

building,

thus

preparing the

surveys of Mr, Koldewey,


of March.

The digging

entirely completed,

were employed
the tombs.

warded

at

all

who

way

for the further

arrived at Assos on the ist

at the

the men,

Atrium being by

now

that time

over forty in number,

in the vicinity of the market-place

and among

In both of these fields the work was richly rethe Stoa, Heroon, and Greek Bath, by finding

inscriptions and architectural fragments which

went

far to-

wards solving the various problems of arrangement and construction presented

by these
^

edifices

Report, p. 20.

in the Necropolis,

by

INVESTIGATIONS

AT

ASSOS,

18S3.

25

the discovery of the finest figurini, vases, and coins obtained

The 24th

during the entire course of the excavations.

March was a day

good fortune.

of

great age were found,

Several sarcophagi of

among them No.

87, containing a

num-

Six hours digging on that day resulted

ber of archaic figures.


in

of

more valuable discoveries than had been made

in this lo-

cality for half a year.

Mr. Bacon having returned to the


the

workmen

engaged

were, during this last

in the Street of

was maintained
of April,

when

of the undertaking,

The number

Tombs.

maximum

at the

site early in April, all

month

of laborers

of forty-five until the 24th

the Easter holidays caused the usual break in

the ranks of the Greeks.

The minor Greek

festivals

were

not permitted to interfere with the work, which was prosecuted with the utmost diligence,

all

the

men being employed

on Sundays whenever showers had caused any considerable


interruption during the week.
Finally,

a close.

on the

ist of

May, the excavations were brought

Throughout the ancient

had seemed advisable

to

city,

every point which

expose had been freed from earth,

excepting only one small corner, about seven by

five

to
it

meters, at

ramp which ascends


had the work been calculated

the western end of the Stoa, beneath the

So closely

to the terrace above.

that forty-eight hours


spot.

But the fear

more would have

of giving the

sufficed to clear this

Turkish

officials

even the

slightest pretext for delaying the division of the objects dis-

covered, or perhaps even for refusing the grant of those to

which the promoters of the undertaking were entitled by the


terms of the agreement, prevented any removal of the earth
after the expiration of the trade,

there had

although

for

some time

been no attempt whatever on the part of the

Turkish government to keep track of the movements of the


explorers.

Thus

it is

not known, and in

all

probability never

ARCH^OLOGICAL INSTITUTE.

26
will

be known, what treasures may

heap

left for

the last because of

Early
tion

in

its

the year the Turkish Ministry of Public Instruc-

matter for some

tions

six

weeks

send an agent

who should

The

delay of a decision in

after the close of the excava-

was more than compensated

for

by the excellence

of

Demetrios Bey Baltazzi, a gentle-

the official appointment.

many services to classical archaeology


Levant, was named as commissioner, and, during two

man who
the

visits

to

the prescribed division and allotment of the antiques

discovered by the expedition. ^

in

concealed beneath that

comparative unimportance.

had been formally requested

make
this

lie

of debris near the entrance to the ancient market-place,

has rendered

to Assos,

from the 27th

from

the

June

to the

of

i6th to the 2ist of June, and

2d of July,

effected

a settle-

ment entirely just and satisfactory to both parties.


The trade under which the excavations had been undertaken was framed in accordance with the laws concerning
antiquities promulgated by the Porte in 1874.^
In regard to
the final division these laws determined that one third of the
objects discovered should be granted to the

where they are found, and one

owner

of the land

third to the finder, while the

remaining third should become the property of the Turkish


government.^

At Assos

the entire extent of the ancient city,

1 Article XXIX. of the Turkish laws relative to antiques, referred to in the


following note, determines that the excavators and the Ministry of Public In-

struction shall each appoint an expert to estimate the value of the indivisible

objects discovered, and to effect a division of them,

provision

being

made

that a third shall be called as umpire in case of disagreement.


2

The

laws on antiques, promulgated Sefer 20, 1291, are given by Aristarchi


troisieme division, Constantinople, 1875,
iii.,

Bey, Legislation Ottomane, vol.


pp. 161-167.
3 " Article

Toute antiquite non decouverte (gisant sous sol), dans


III.
quelque endroit qu'elle se trouve, appartient au gouvernement. Quant aux
antiquites trouvdes par ceux qui effectueraient des fouilles par autorisation, un
tiers appartiendra au gouvernement, un autre tiers au trouveur et le reste
au proprietaire du terrain

oil les

antiquites ont ete trouvees.

Si le trouveur a

INVESTIGATIONS
within the walls,

is

vakouf}

tenance of the mosques,


in question

of

is

AT ASSOS,

a domain

and

2J

set apart for the

main-

hence, in so far as the point

concerned, a domain of the state.

Two

thirds

the objects discovered were therefore exacted by the

all

Ministry of Public Instruction.

The

only private enclosure

where excavations had been made was the


the west of the Street of
dion,

1883.

Tombs, the

level field lying to

site of the

ancient Sta-

which had been recently reclaimed and sown with wheat.

The owner of this ground had sold to the expedition, for the
sum of three Turkish pounds, the right to dig in certain parts
of the field, ceding also that portion of the antique objects

which would by law


here, apart from

to

fall

But the discoveries


mausoleum with the barrel-

his share.

the important

vaulted ceiling, were only some half-dozen vessels of coarse


pottery, not of sufficient value to render

it

advisable to enter

a protest against a general division on the terms before

men-

tioned.

Attention was

devoted to the coins,

first

three thousand had been discovered.

once

Those

set aside for consideration with the

cious metal.

of which nearly
of gold

ornaments

were at
of pre-

Sixty of the coins of silver and bronze were of

especial interest on account of the positions in

had been discovered


thereby determined

in sarcophagi, the

which they

ages of which were

under walls and pavements, thus

able to subsequent dates

and

in

refer-

accumulations of Byzantine

and mediaeval debris, attesting the overthrow and desertion


of the various sites.

As

these coins were of greater impor-

trouve ses antiquites dans sa propriete, les deux tiers seront a


au gouvernement."

lui

et le reste

1 The
laws governing vakouf property are given in the Legislation Ottomane, before quoted, vol. i., section deuxieme, Constantinople, 1873, PP- 241There is still no better popular explanation of this peculiar institution of
249.

the Turks than that given by Mouradja d'Ohsson (Ignace de), Tableau general
de I' Empij-e Othoman, Paris, 17S8-1824.

ARCH^OLOGICAL INSTITUTE.

28

tance to the expedition than to the Turkish Museum, the

bulk of them was kindly granted to the investigators by

who chose

Baltazzi Bey,

as an equivalent one

hundred and

He

twenty of the best preserved specimens remaining.


not think

did

worth while to sort and count the oxidized and

it

defaced coins, but weighed out the twenty-five pounds or more

with scales borrowed from the village bakhal, allotting alternately one oke to the American, and two to the Turkish share.

This method of division was,

somewhat crude but,


it was impossible to
Nine hundred and eight coins thus
it is

true,

as the pieces were well mixed together,

complain of

it

as unjust.

became the property of the expedition, two hundred and fiftyseven of which were of numismatical interest, and had been

Among

identified without the aid of a specialist.

no

self,

many

these were

one hundred and twenty-two coins of Assos

less than

of

them

of silver

the finest specimens, and

all

it-

the

imperial types minted by the city, being obtained in exchange


for certain coins of the Diadochi
intrinsic,

On
liefs.

but of

less scientific value.

the 28th of June a division was


In this

important matter

a satisfactory conclusion, and


sion,

and Byzantines, of greater

and by giving up

to

it

it

made

was only

the

of the

was more
Porte

all

temple

difficult to

after

much

re-

reach

persua-

the fragments of

bronze sculptures discovered during the excavations, that the


expedition secured the two finest blocks of the epistyle,

namely, the Herakles with the human-legged centaurs, and


the two heraldic sphinxes from the eastern front of the building, superior to all the others in

workmanship and

of better

preservation.

An

was made in regard to


The commissioner considered the value

especial arrangement

scriptions.

the inof the

bronze tablet, with the oath of the Assians to Caligula, as


equal to twice that of

all

the inscribed stones together, and

INVESTIGATIONS
could not be prevailed upon to

AT ASSOS,
make any

29

1883.

allotment by which

the Porte would be obliged to relinquish this treasure.

The

possession of the tablet was, indeed, greatly to be desired,


as

it is

one of the largest and best preserved among the few

Greek

bronze inscriptions remaining from

though

its

entirely lost through

two years' exposure

its

Turkish custom-house, where

by the

Maimouri,

first

its

make it in this respect


any museum of antiques.

the division of

all

also

one

to the air in the

still

so

striking

an acquisition to be prized

Nevertheless,

it

was

felt that, in

the seventy-four inscriptions discovered at

Assos, those cut in stone which

from the

Al-

had been placed under seal

it

appearance was

as to
in

antiquity.

patina of brilliant green and blue had been almost

it

was possible

remove

to

formed, in essential value, decidedly more than

site

In historical interest, for instance, the bronze tab-

third.

let is certainly

not equal to either the inventory of the great

temple, the dedicatory inscriptions of the Bath, or the epitaph


of Hellanikos

inscriptions,

Porte,

and

Arlegilla.

The proposed

division of the

by which the bronze alone was taken by the

was therefore accepted without demur.

The marble

sculptures, figurini, pottery, glass, and miscel-

laneous objects were divided, class by

each into three approximately equal

class,
lots,

by Baltazzi Bey,

the choice of one

of these being allowed to the investigators,

and indeed favorable

to the choosers.

been much more trying


finders to

make

if

It

^his was

fair,

would certainly have

the commissioner had required the

the division, and leave to him the selection

of two of the thirds.

Baltazzi

Bey, however, arranged his

lots

with surprising equality, so that the advantage of the

first

choice was not so great as might have been expected.

He was

uniformly obliging in putting into the same share

objects which in any wise belonged together,

was

whenever

possible without disturbing the relative values,

this

ARCHAEOLOGICAL INSTITUTE.

30

The

attention

of

explorers

the

was

invariably

directed

towards the acquirement, in so far as was possible in this


small fraction of the whole, of representative types,
entific rather

than material value.

and

tiques, as of the coins

reliefs,

of

sci-

Thus, of the minor an-

the Americans secured the

most interesting specimens, although

not quite one third

become evident by a compariwhich


have
been removed to America
the
objects
son of
Both
with those remaining in the possession of the Turks.
of

them

will

number.

be described

former

will

them

to

in

The

in

This

will

in detail in

be referred

to

the Boston

the subsequent pages, and the

according to the numbers attached

Museum

of

Fine Arts.

division could not be otherwise than a painful task to

Series of figurini, glass, vessels of terra-cotta,

the explorers.

and many minor objects

illustrative of the industries of the

even the
together within one grave, had often
city during various ages,

trifling

to

memorials buried
be separated, not-

withstanding the obliging readiness of Baltazzi Bey to comply

was with sadness that two thirds of the


antiques which had been acquired by such long and hard
labor, and had come to be viewed almost with a feeling of personal attachment, through the familiarity of close study, were
with our wishes.

It

given up to the unheeded corners of a Levantine museum.


It is

but ju^ that attention should be called to the

exceptional,

Turks,

if

not unparalleled in

fact,

dealings of this kind with the

that not the smallest object, not a single coin or sherd

was kept back from the division by the explorers.


The instructions given in this respect by the executive committee of the Archaeological Institute had been explicit, and
of pottery,

were carried out by their agents with scrupulous exactness.^


should be stated that these instructions were in conformity to the obligaby the Institute in the acceptance of the iradi ; the executive
committee being bound in honor, no less than in morals, to issue them.
1

It

tions entered into

INVESTIGATIONS AT ASSOS,

18S3.

only antique removed from the site during the progress


a gold coin found during the first year, and
of the work

The

submitted to an eminent American numismatist for determination

was

ultimately

the

returned by mail before


to the share of the Turks,

fell

and

division,

who have

it

now

Only those experienced in Oriental methods


of dealing can fully understand what this means. An entirely
different procedure would have been quite in accordance with
and
the accepted laws of human intercourse in the Levant

in

possession.

this being naturally taken for granted by the authorities, it


was utterly impossible to convince them that the usual pro-

were

testations of fair dealing

in this case literally true.

certain license of appropriation enters into the calculations of


all

Turkish business

and, as in

of individual

most instances

deviation from established usages, the consciousness of abso-

was here purchased

lute rectitude

disadvantages.

In

pride by those

who

work

but, as a

jiistitia of

On

itself,

at

the expense of great

may be regarded with


by those who carried out, the

this position

planned, as

moral lesson to the Turkish

the Archaeological Institute was

the other hand,

it

is

certain that

official,

the Jiai

certainly futile.

some few antiques


workmen, in

were stolen from the excavations by the Greek


spite of

The men were narrowly watched

precautions.

all

during the work, and received,

in

addition to their regular

wages, gratuities for such small objects as they brought to


the explorers. The obliterated and less valuable coins were collected every Saturday night,

were bought

in

at a fixed

it is

true, to their

still

rather

worth

when

the pay-roll was called, and

scale of prices,

for the

purposes of investigation, but

more than they would have fetched

bakhals, or to the travelling Jews,

modest speculation
of figurini.

incommensurate,

No

in

ancient

who

if

sold to the

usually carry on

gems, coins, and fragments

instance of an antique being secret-ed

by

ARCH^OLOGICAL INSTITUTE.

32
a Turkish

but the

workman was

detected during the three years

Greeks were often tale-bearers, as well

thieves, and exposed various acts

of dishonesty

as

petty

on the part

of their fellows.

The lower classes of Mytilene, from which the greater


number of the Greek laborers were recruited, are notorious
throughout the East

for their

rhyme

the well-known

sharp practices, as attested by

'AOrjvaiOL Kol Qrj/SoLOi

Kat KaKol ^IvtlXtjvolol,


"AXAu Xeyovv to fSpaBv
Kt

cikXa Kd[xvovv to ra^i'.

Indeed, the inhabitants of this island have often been stig-

matized as

among

their ancestors

Greeks.

the most unprincipled of the modern,^ as

were among the most depraved of the ancient

Hence

could be no surprise that pilfering was

it

attempted, notwithstanding every incentive to

and that some small

Whether any

the strictest surveillance.

were

of real

fair

dealing,

thefts passed without detection, despite


of the objects taken

importance cannot, of course, be known.

at least impossible to steal

It

was

from the trenches anything which

could not have been concealed upon the person during the
day's work, and the

most important investigations

those

concerning the architecture, sculpture, and epigraphy of the


ancient city

cannot

have suffered in the

fragments of terra-cotta

from Assos, were

in

figurini, said to

the

hand

of

slightest.

Some

have been brought

a dealer at Smyrna, in

May, 1883, and were offered for sale at an exorbitant price.


But Professor Ramsay, from whom this information is derived, states that these were of little value, and certainly
1

Compare the remarks of Finlay (George), A History 0/ Greece /rem its


Romans to ... A. D. 1S64. (New Edition.) Oxford, 1877,

quest by the

vol. V. p. 60.

Conetc.,

AT ASSOS,

INVESTIGATIONS
much
sos

inferior to

itself.

it

and

prove

this

a mere hy-

is

any of the better images, which have since

come into European and American


by way of Smyrna, were originally
tions at Assos,

it

collections of antiques

stolen from the excava-

will in that case also

that the science of archaeology


rectly,

33

the specimens which he had seen at As-

Even should

pothesisthat

1SS3.

not the

is

be borne in mind
because indi-

less,

indebted for these contributions to the promoters of

the undertaking.
to dig

The only unauthorized attempt

upon the

site,

dur-

ing the two years designated by the imd/, was made by one
of the Greek valonea merchants, in a spot where nothing

but sherds of pottery and broken terra-cotta figures could be


found: a terrace adjoining the lower fortifications, filled in

with debris during antiquity. Some few basketfuls of earth


were here removed. This should not be considered as any
intentional infringement of the rights of the expedition, but

rather as a continuation of the desultory digging which had

been carried on upon the ancient

many

site for

years.

In

this instance the objects discovered, ultimately handed over to


but before the comthe expedition, were of no great value
;

mencement

of the systematic excavations the treasure-seekers

had often been more successful.


while digging

among

In 1878 a

Turk

of

Behram,

the ruins, discovered twelve silver spoons

These he carried across to the island of


of curious shape.
An attempt was made to
Mytilene, and sold in Skamnia.
find them, but they had passed from hand to hand and had
finally

been melted up

images

of sheet silver

for the

manufacture of the hideous

suspended as votive offerings at the

shrines of popular Greek

saints,

just

as

certain of

the

prehistoric gold ornaments stolen from Hissarlik were trans-

formed into modern jewelry.

So firmly did the inhabitants


3

of

Behram

believe an endless

ARCH^OLOGICAL INSTITUTE.

34

among

wealth of precious metals to be hidden


of the ancient city, that

men

of the village would

had seen

dream a

in a

now and then one

every

come

of the old

how he

to tell the explorers

treasure, generally a pot of

buried

beneath a certain spot,

locality

if,

when

the foundations

offering

money,

indicate

to

the

the prize was found, he might be allowed a

share.

This belief

ment

as

hopes.

in

One

aspects.

hidden treasure was not without

of the subordinate officials sent

ludicrous

Maimouri long entertained the most extravagant


would jump into the trenches whenever a sar-

He

cophagus was unearthed,


confidently expected
to the

its

by the govern-

it

in order to seize the gold

which he

This became troublesome

to contain.

workmen, who could hardly be

seriously reproved for

causing him the transient delight once excited by the pre-

tended discovery of a heavy and shining bowl of yellow


metal.

No damage was

done by the excavations

to

any property,

public or private, for which the expedition could have been

held responsible according


articles of the
flict

Turkish code.

to

the

And

fifteenth

and sixteenth

no case was there con-

in

or disagreement with owners of fields or sheep-folds,

jealous as the villagers naturally were concerning any disturb-

ance of their enclosures.

By
up

the night of the 30th of June, Baltazzi

all

the movable antiques which had

Bey had packed

fallen to the share of

the government, and had deposited them under seal in the

magazine
It

of the

customs

official of

the port.

had been hoped that the Turks would be readily induced,

after the division, to sell the greater part of the antiquities

which had thus become their property,

more especially the

remaining blocks of the temple epistyle.

seum

of Fine Arts

had

set aside the

sum

The Boston Muof $2,000 for this

INVESTIGATIONS
purpose.

But

it

AT

ASSOS,

1SS3.

35

soon appeared that the Ministry of Public

Instruction, in extreme pursuance of the principles adopted

shortly before, had determined to forbid the sale of antiques


to foreigners, as well as to prohibit the exportation of

from

ancient art

of

the

works

proposals

purchase of the Turkish share of the objects

relative to the

discovered at Assos would be entertained

ment

No

Ottoman Empire.

all

and, as the govern-

had exercised the option allowed by the

the code then in force, ^ and had divided

all

fifth article of

the antiques eu

nature, and not en valeur, no further claims could be made.

The sum voted by

the Boston

Museum

was

of Fine Arts

there-

fore unused.

An anonymous

writer in the

New York

Nation

has since

urged the Archaeological Institute to undertake legal proceedings against the Turkish Ministry of Public Instruction,

on the ground

that, as

he asserts, the law of 1874 "enacts

that indivisible sculptures shall not be divided,

and given up

to the

finder in

Unfortunately, there

value."

to bear out this assertion.

is

exchange

nothing in the law

The

finders

but valued

for their estimated


in

question

were by no means

thus favored with the certainty of ultimate possession, the


decision as to any proposed purchase being

option of the government.

pealed before the division


at present in force,

1884,

left

wholly to the

The code of 1874 had been rewas made at Assos, and the laws

although not promulgated until February,

had already been determined upon.

This of course

altered nothing in regard to the terms and privileges of the

trade ; but the commissioner had been instructed to negotiate

no

sale,

and

to allow the exportation of

no antiques excepting

those which the government was pledged to give up according


^

" Article V.

La

repartition des antiques se fera, selon la

gouvernement, en nature ou en valeur."


2

The Nation,

New

York, Nov.

13, 1884.

demande du

ARCH^OLOGICAL INSTITUTE.

36

agreement.

to the original

policy that Baltazzi

It

was

conformity with this

in

Bey had made out

the lots in kind.

all

1874 had actually contained any such


which the well-meaning critic in the
upon
clause as that

Even

if

the law of

Nation would have based a legal claim,

it

would, under these

circumstances, have been impossible to persuade the Turks


that the sculptures from the epistyle of the temple were to

be considered as

indivisible,

they being well

aware that

at

that very time a considerable part of this disconnected series


of representations

was

in Paris.

In conformity with the thirty-second article of the code,


detailed lists of the antiques belonging to the

were made out and submitted

Behram and
Vilayet.

to

of the Dardanelles,

American share

the customs

the

of

officials

chief station of the

Export duties are levied by the Turks even upon

fragments of ancient works of

art,

eight per cent being as-

sessed on the assumed value, the determination of which

is,

of course, in such a case, altogether arbitrary.

While awaiting a decision

and the issue of

in this matter,

the teskere which should permit the removal of the antiques

belonging to the expedition,

all

the objects were packed in

the inscriptions and smaller stones being


wooden cases
wrapped in hay, the vases and figurini in a fine dried seaweed. The greatest care was required in preparing the glass
vessels for the long transit, as many of these had become
:

exceedingly fragile through the flaky oxidization of two thou-

sand years, the iridescent material being

most

as thin as paper.

pitcher,

now

own weight

in Boston,

while

still

in

some places

al-

Indeed the side of one delicate glass

was found
in

the

to

have

fallen in

sarcophagus,

so

from

that

its

this

exceptionally fine specimen was excluded from the division

by the commissioner, on the ground that it could not possibly


be removed from the site. In packing things of this kind an

INVESTIGATIONS

AT

ASSOS,

1SS3.

2)7

expedient was adopted which proved entirely successful, and

may be recommended

Fleeces of cotton

for similar cases.

were lightly held over the object while a spray of thin mucilage was blown upon them through an atomizer.
After this
had been allowed

to dry, further layers of cotton

were wrapped

about the whole, which was then wound around with thread

and dipped into thick


proved

packed

perfect

glue.

protection,

in a separate

box that

The

firm

especially

shell

as

ball

was

might not be exposed

it

In opening, the outer layers were

pressure.

thus formed

each

cut

shears and the innermost fleeces slightly moistened.

off

to

with

Three

ancient skulls, which seemed so ready to crumble into dust


that even their removal from the Necropolis to the magazine
at the port

was

at first

thus prepared, and

thought hardly possible, were also

when placed

in

the hands of that emi-

nent specialist, Dr. Virchow of Berlin, they were

in precisely

when taken from the tombs. In all the


long voyages and many transshipments between Behram and
the same state as

Boston the only one of the antiques injured was a vase which
separated in the lines of an old crack.

Every

article

belonging

to

the

American share being

packed, the export duties paid, and the permission to remove


the cases having been received by telegraph from the customs
officials

of the Dardanelles, the writer left

14th of July, 1883.


for

some time

the site on the


Mr. Bacon and Mr. Koldeway remained

longer, in order to complete the detailed surveys

of the Necropolis

and Agora.

were subsequently raised by the Turkish offiregard to the removal of the architectural fragments.

Difficulties
cials in

These consisted

of a complete order from the great temple of


the Acropolis, the stump of an archaic column, and specimens
of various mouldings from the Street of Tombs, capitals from

the Stoa and Bath, and portions of the two chief mosaics of

ARCHyEOLOGICAL INSTITUTE.

38

the lower town.

but had

all

been

They had not been included


left

to the

in the division,

Americans, as Baltazzi Bey did

not consider them of sufficient value to the Imperial


to warrant the expense of carrying

them

Museum

to Constantinople.

After the local officers of the customs had permitted the ship-

ment

of the other objects, they found a pretext for interference,

and stubbornly refused to allow the removal


because they had not been

specified

These stones had been admitted to be


Turks, and their retention by the customs

of those blocks,

the

in

division

lists.

of no value to the
officials

was simply

an act of obstructiveness,

an abus de pouvoir, as the before-

mentioned writer

in the

Nation has justly characterized

The commissioner

could not again be

summoned

it.

to the site

and, in spite of the repeated efforts of the Archaeological


Institute through the

American Legation

at Constantinople,

nothing has since been obtained but promises from the Director of the Imperial

Instruction, to

The

total

Museum, and from the Minister

whom

the ultimate decision

is

cost of the investigations at Assos,

every expenditure at

all

of Public

referable.

including

connected with the undertaking, from

November, 1880, until May, 1884, was $19,121.16. Of this


no small
sum $3,344.53 was directly spent in earth-work,

proportion of the whole, in view of the remoteness of the site

from the home of the explorers, and the fact that attention

was

at all times directed rather to a scientific investigation

of the

monuments of the ancient city, than to a mere sifting


mounds of rubbish in the hope of discovering anThe remainder was devoted to the purchase of the

of great
tiques.

household and excavating


party of explorers

outfit, to

upon the

site,

the maintenance of the


to

the transportation of

the chattels of the expedition and the antiques discovered

by

it,

to the travelling

expenses and the salaries of the agents

of the Institute, and, finally, to the expensive official relations

INVESTIGATIONS

The

entire

1883.

39

work carried on under Turkish


outlay, although a heavy tax upon

inseparably connected with


jurisdiction.

AT ASSOS,

all

the resources of a newly formed society of private individuals

cannot be thought excessive, considering the extended and

complex nature

of the task:
'EfcaTov Se tc 8ovpa6' d/Aa^s.^

has pleased the executive committee of the Institute to

It

express

its

entire satisfaction with the detailed accounts ren-

dered by the agents employed in this work, and to praise their


constant economy, upon the exercise of which the success of
the expedition was in great measure dependent.'-^
^

Used

Theaitetos,

proverbially by Plesiod,

XLII.

Works and Days,

\ Archaological Institute of America.


Committee, p. 26.

456,

and quoted by Plato,

17.

Cambridge, 1S84.

Fifth

Annual Report of

the Executive

CHAPTER

II.

ACROPOLIS AND TEMPLE.

THE one

volcanic crater which forms the AcropoUs of Assos

is

Troad.

most striking natural features

of the

the

of

Rising precipitously on a narrow strip of land be-

tween the sea and the river

the highest point

valley, it is

between the mountain of Qozlou-dagh (Gargara), ten kilometers to the east,

and the great plateau above Polymedion, an

equal distance to the west.

To

hence very marked.


port

Its

topographical isolation

in all the wonderfully picturesque lands inhabited

is

repeat the words of the First Re-

by

the Greeks, no site was more majestic or more beautiful than


that of Assos.
that,

The

cliffs

upon the seaward side are so

standing on the Acropolis, one can look

holds of the vessels moored in the


so high, that the

On

morning

summit

is

little

into the

port beneath,

at times in or

in early spring, while

down

steep,

and

above the clouds.

drops of rain were

fall-

ing at the port, the writer has climbed through a thick bank
of vapor, hanging between the

Agora and

the Acropolis, and

has found the sky blue overhead, and the ruins of the temple
lighted up

The

by the

finest

first

rays of the sun.

views of the Acropolis are to be had from the

southeast and the southwest, a mile or more out to sea.

It

was, without doubt, partly on this account that the temple

which crowned the height was placed close


edge of the summit.

mark from every

The

to the

southern

building must have formed a land-

part of the Gulf of Adramyttion and the

INVESTIGATIONS
Strait of Mytilene

AT

ASSOS,

thrown into

The

cella wall.

41

the sailor nearing the port could grad-

ually distinguish the quiet lines of its


lature,

188S.

relief

by

columns and entab-

dark shadows upon the

their

low, mediaeval towers

now surmounting

the

Acropolis can be distinctly seen from the fortress of Molivo,

twenty kilometers away, although at

Assos

is

this distance the hill of

But

not outlined against the sky.

dependent

in great

measure upon

its full

grandeur,

isolation, is felt

its

when

the observer stands upon the heights separating the river


valley from the sea,

for instance, at the village of Bourgas,

upon the west-northwest, or on the road

to

Sonoba, to the

east of the Acropolis.

On

more gradual, but even here the


cliffs are often twenty meters high.
Towards the sea, the
grade is almost one in two towards the river plain, it is one
the north, the slope

is

While, on the one

in four.

side,

the crater and

its

surround-

ing dikes have been scarped by the action of the waves, on


the other the great ravine between the two ranges has been
filled

in

with the detritus brought by the Satnioeis from the

heights of Ida since the close of the tertiary period, so that


the stream

now

flows at an elevation of one hundred meters

above the level of the sea.

The

earliest inhabitants of the land

cannot have

failed to

take advantage of the protection afforded by this great natural stronghold.

Geological revolutions, which in one epoch

plunged the entire range beneath the

The

sea,

had truncated the

summit thus formed could


easily be rendered impregnable by the erection of walls at
those few points where the foot of man could make the ascent.
The limited circuit could be defended by a small number of besieged against the attacks of an army and yet the
enclosed area was of sufficient extent to accommodate a
cone of the

crater.

level

considerable garrison.

ARCH^OLOGICAL INSTITUTE.

42

As

to give every possible advantage, nature not only-

if

surrounded the rocky heights with the most

fertile

fields,

but created, within the very confines of the citadel, an inexlarge cavity in the rock
haustible reservoir of water.^

which forms the northern terrace


on the part of

effort

man

when no

present day,

of the Acropolis retains a

midsummer, without the

provision of water until

to increase

furrows, even, are

made

inhabitants of

when

taken to keep

is

Behram can

ing upon

in the

rendered accessible.

its sides,

it

unde-

season as July.

the revetement of the deep fissure at the

descend to the water with

The

and

depth, measured from the floor of the eastern

to the

width of the cistern

spring of the vault, 7.5 m.

women

date from the Middle

their jars, all

Ages.

total

first

The masonry now remain-

bottom, and the flight of steps by which the Turkish

its

no

fetch their drinking

not possible to say at what period this cavern was

enlarged and

the

along the earth to carry the winter

water from this cistern as late


It is

At

buildings stand upon the Acropolis

torrents into the well, and no care


filed,

slightest

supply.

roofs the rainfall could be collected,

from whose

the

its

6.8 m., its length 8 m.,

is

chamber

longitudinal division

wall has been built, and the two subterranean chambers thus

formed have been covered with rude barrel-vaults, the imposts of which are nearly on a level with the ground.

enormous

reservoirs, once

filled,

to supply, for an entire year, the needs of a garrison of

than six hundred men.

These

would contain enough water

Replenished by every

rain,

more

they must

have been capable of providing drinking water for all the


twelve or fifteen thousand inhabitants of ancient Assos, at
the time of
It
1

is

its

plain

Shown on

" Cisterns."

greatest extent.

that

the

walls of defence

erected upon the

the plan of the Acropolis, given in the First Report, Plate

2,

as

INVESTIGATIONS

AT

ASSOS,

1883.

43

Acropolis were in great part razed to the ground after every

Their stones have gone to increase the

successful attack.

great slopes of rubbish which have accumulated at the foot of

the

cliffs.

Nowhere, throughout

their entire extent,

have the

fortifications of the

town been so thoroughly demolished as

upon

The

this

height.

investigations concerning the mili-

tary architecture of ancient Assos

ing walls supply so


derive no

of which

many examples

materials whatever

the surround-

in perfect preservation

from

this

strongest retreat.

Only here and there are vestiges of polygonal masonry

to

be

seen, these generally forming low retaining walls in clefts of

the

native

The

rock.

only courses of accurately squared

stones in position are those beneath the huge tower of rubble,

erected in the Middle Ages, which, with the other mediaeval


fortifications of the Acropolis, will

quent chapter.

Even

be described in a subse-

the latest and rudest ramparts, con-

sisting of small stones set in thick mortar, have, with few

exceptions, been levelled to the ground by the Turks, who,

of the successive inhabitants of the place,

seem

only ones that have done nothing to increase

to

be the

defensive

its

strength.

Evidences of the attacks made upon the fortress are seen


in a

number

of

human

skeletons, found buried in the rubbish

which had accumulated around the foundations


ing wall upon the southeast.
of

besiegers

who had

fallen

of the enclos-

These must be the remains


in

assault,

and whose bodies,

covered with the stones and earth thrown over the brink, had
not received the rites of sepulture.

Fragments

of at least

three skulls were here brought to light, but the bones crum-

bled at a touch, and could not be removed from the site for
osteological study.
riors

had

perished

The
was,

period at which one of these war-

however, determined, with some

degree of certainty, by the character of the weapons found

ARCHAEOLOGICAL INSTITUTE.

44

near the bones, to be that of the conquest of Assos by the


Persians in the sixth century before Christ.

In the well-

defined stratum in which these remains were buried, three

bronze arrow-heads were discovered, one in particular lying

The arrow-heads found

close to the skull.

depth were

at this

type,

now preserved

last referred to,

Boston (Museum, No. A.

They

resemble

50,

among

re-

age of the Achae-

mains referable

to the

menidae,^ and

those picked up by Gell

upon the
I.

true,

to

Marathon.^

field of

Bronze

Archaic Bronze
Arrow-head.
From the Acropolis.

Fig. i).

every particular the

in

arrow-heads found by Ouseley

Fig.

same

and of a shape exemplified by the

specimen
in

of precisely the

all

arrow-heads

be employed

period of antiquity.

continued,

until a

This

is

later

proved by

is

tombs of Greek warriors

their presence in the

much

it

at Kertsch,

in Southern Russia, dating from the age of Alexander the

Great,^ as well as

among

Naukratis* and

the Altis of Olympia.

in

the remains brought to light at

But the arrow-

heads in use subsequent to the Persian wars are readily to

be distinguished from such archaic ones as those found in


the lower stratum.

Specimens

of a three-bladed variety

the slopes of the Acropolis of Assos.

(Museum, No. A.
1

49, Fig. 2).

Its

were also found upon

One

is

now

in

Boston

form illustrates the term

Ouseley (Sir William), Travels in various Countries of the East, more


London, 1819-23, vol. ii. PI. 39.
Gell is quoted, and the arrow-head in question is engraved in the work

especially Persia,
2

referred to in the preceding note.

Academie de

St.

St.

Petersbourg, Compte-rendu de la Comtnission Imperiale,

Petersbourg, 1S76, Plate


*

The

II.

objects in question were

antiques from Naukratis, held

examined by the writer at the exhibition


Oxford Mansions, London, 1S85.

in the

of

INVESTIGATIONS
rpiy\coxi<i, applied to

AT

ASSOS,

an arrow by Ilomcr.^

1SS3.

45

These heads

of tri-

angular section seem, like the ones before

mentioned, to have been imitated from a


Persian type.^

Indeed, the Greeks ap-

pear to have derived

edge of the

much

fittings of

well-equipped

of their knowl-

archery from the

bowmen who made up

the

main force of an Oriental army.^ Witness


their " Parthian" and " Scythian" bows.

Among

the

weapons found around

the walls of the citadel are also an iron

spear-head with a long and thin blade of


fine

workmanship, now much corroded

(Museum, No. M.

Fig.

598,

a,

d),

and a

heavy double-headed axe or mattock,

2.

Bronze Arrow-head.
From the Acropolis.

excellently

preserved, which, notwithstanding its modern


appearance, was found in a situation and at a depth proving

Fig.

3.

From
at

it,

all

events,

(Museum, No. M.

to

Iron Mattock.
the Acropolis.

have

antedated

the

Middle

Ages

603, Fig. 3).

1 I/iad, V.
Elsewhere in the I/md the fittings of arrows are
393, XI. 507.
especially referred to as of bronze (e. g. XIII. 650, C62).

Compare the ancient Persian arrow-heads given by James P. Morier, A SecJourney through Persia, Armenia, and Asia Minor, etc., London, 1818.
8 Almost all the troops which formed the army
of Xerxes were armed with
the bow.
Compare Herodotos, VII. 61-80. Bronze arrow-heads of this threebladed kind are, however, said to be found in Greece upon every spot where
^

ojid

a battle

is

known

to

have been fought.

See Dodwell,

graphical Tour through Greece, London, 1819, vol.

ii.

p. 160.

Classical

and

Topo-

ARCHAEOLOGICAL hWSTITUTE.

46

Within the enclosure

of the Acropolis

no such arms or

implements were found, and the only human remains were


the crumbling bones of one individual, contained in a cist
neatly constructed of large

tiles,

which had been buried

in

the deep earth at the north of the temple, 80 cm. above the
native rock.

The

coins picked up in the trenches furnished specimens of

the mintage of every age, from the time

was

first

man

employed

in the

Troad

when coined money

until the

advent of the Otto-

Remarkable among these were a

Turks.

fine electron

of Michael VIII., Palaiologos,i and a silver coin of the Ve-

netian Dandolo, one of the few memorials of the occupation


of Assos by the Latins after the

fifth

Crusade.

It appears that no buildings have ever been erected by the

Turks within the walls


there can be

little

On

of the Acropolis.

doubt that

many

the other hand,

small dwellings and store-

houses did stand within the enclosure during the two centuries

which intervened between the

and the Ottoman occupation.


citadel

was covered

at this

Seldjukian conquest

first

The southern

which were embedded small stones and

in

terrace of the

period by a pavement
bits

of cement,
of pottery.

This pavement was on a level with the stylobate of the


temple, the massive blocks of which served as the foundations for a confused

group of hovels and magazines.

At

the

northeast, similar structures of rubble walls without mortar

were
est

built directly

upon the native rock

and near the high-

peak were discovered the lower courses of an apse,

part of a small sanctuary, such as the Byzantine Christians

erected upon

many

of the neighboring eminences.

All these

structures were enclosed by rude ramparts, consisting, on the

west,
^

almost entirely of blocks of the entablature of the

Compare

First Report, p. 32.

share of the Turks.

In the

official

division this coin

fell to

the

INVESTIGATIONS

AT

ASSOS,

1SS3.

temple, and, on the south, of a long array of

At

placed on the sides of their abaci.

47
its

capitals,

the east, ample de-

fence had been provided by somewhat earlier mediaeval walls,


the stones of which were set in mortar; and, at the north,
the precipitous

rising

cliffs

from the lowest

to the highest

crowned only by a low parapet.


enclosure, at the west,^ was flanked by

terrace of the Acropolis were

The entrance

to this

upright beams of the inner epistyle of the temple.

Most of

the reliefs discovered by the expedition had been built into

these walls

the lion and boar, the lion and hind, and the

entire metope, standing at the west, the sphinxes from the

eastern front of the temple at the east, and the Herakles and

Centaurs

at the south.

The cemetery
the

of the Byzantine garrison

Here

Acropolis at the northeast.

was

just outside

the bodies of the

defenders were buried in the shallow earth, without being

enclosed in coffers of any kind.

number

of these graves

were opened, but in them were found no coins or weapons


by which their age could be determined.
Still, we may
venture to suppose that the most recent and hasty ramparts

were erected

at

the beginning of the

under Machrames, the

whose pathetic

last

fourteenth

defender of the

history will be recounted

in

century,

Greek town,
a subsequent

chapter.

How

completely the temple had by this time been destroyed,

even the pavement of the temenos being torn up and washed


away, became evident from the position of one of the capitals,

which was found lying with


below the
It is

rise of the

now

impossible to determine the original plan, or even

the extent, of the upper terrace


size of the boulders
1

abacus nearly half a meter

its

lower step.

which have

but
fallen

it

is

evident, from the

upon the

First Report, Plan of Acropolis, Plate

2,

" Gate."

floor of the

archjEOLOGICAL institute.

48

temple, that the

bank

of earth

must have been

So large were several

able height.

most recent occupants

up the wretched masonry


against them.

was ever the


tiges of

it

building upon the

of the citadel, in

plan of the edifice, did not even


of

of consider-

of these stones, that the

them away, but

roll

dwellings

their

piled

around and

seems improbable that the upper terrace


any important monument, as some ves-

It

site of

would certainly have been found.

overgrown with verdure

in ancient times,

It

was probably

and served as the

Here must have stood the inscribed

peribolos of the fane.

stones relating to the temple, one of which, containing an

inventory of chattels, was brought to light by the excava-

The

tions in the vicinity.^

was

also found

pedestal of another inscription

upon the Acropolis, and, as

mouldings rendered

useless

it

scarcely be supposed

to

as

its

projecting

building stone,

can

it

have been carried up the steep in

later times.

The

stylobate of the temple

is

fully

three meters lower

than the highest point of the rock, and must consequently


have been at least so much below the level of the upper terrace.

From

from the

the sea, the whole structure was visible

river valley, at the north, little

but

more than the

roof

could have been seen.


In view of the ample space provided by the level bed-rock
of the southern terrace,

it

tation of the temple, with

may be
its

conjectured that the orien-

longitudinal axis deviating from

the east, was due to the consideration that this position of

the building permitted the gable of the front, as well as the

long horizontal lines of the side entablature, to be seen from


the chief places of the lower town,

and the direct ascent from the


no doubt that this exceptional
^

the

port.

agora, the theatre,

There can

at least

relation of the plan

Inscriptions of Assos,

No.

III.

to

be
the

INVESTIGATIONS

AT

ASSOS,

1SS3.

49

this intenpoints of the compass should be ascribed either to


a similar
to
or
tion of exhibiting the foreshortened front,

motive of

artistic

composition

namely, the desire to place

in order
the building near the seaward brink of the Acropolis,
should
be visthat as much as possible of the superstructure

from below.

ible

Foundations considerably deeper would

have been necessary beneath the southwest corner, if the


eight and a
front had faced due east, and had not been set
half meters

back upon the summit.

This was, in truth, a fitting site for the temple of the proFor the
tecting goddess, the Virgin Patroness of the town.
fane was not only visible from afar, but was so placed that the
Assians, while offering sacrifices for the welfare of their state
within the sacred enclosure, could look far beyond the fortification walls, the fertile fields of the suburbs, and the port

beneath the

by

sea.

cliff,

On

one

most distant approaches by land and


the view commanded the rugged paths

to the
side,

winding across the ranges of the

interior,

whence,

in the evil

days preceding the establishment of the supremacy of Pergamon, came devastating hordes of Gauls on the other, the
;

deep blue waters of the Lesbian Straits, the great high-road


Even the tiny islets of the Arginousai
of Aeolic commerce.
can be distinguished from the Acropolis of Assos

and

if,

on

the day of that victory so disastrous to the best interests of


Greek culture, the horizon was not veiled by the rack of the

storm which delayed the Spartan attack, and served as an


excuse for the inhumanity of the Athenian admirals, a sharpsighted observer, standing on the steps of the temple, might
have followed the movements of the rival fleets, exulting over
the victory of that power with which his

Demos was

then

allied.

From
was

stereobate to corona, the stone of which the temple

built

was the same

as the native rock


4

upon wliich

it

ARCH^OLOGICAL INSTITUTE.

50
Stood.

The

only other stone employed in connection with

the edifice was a light volcanic tufa, of the

same geological

formation, from which were carved the gargoyles and acroteria of the roof.

The mosaic

of later date, in the interior of

the naos, formed of small cubes of black and white marble,

is

rather to be regarded as a furnishing, than as an integral part


of the structure.

Before entering into a detailed consideration of the plan

and elevation of the temple,

it

be well to give some ac-

will

count of the rock of the Acropolis, the peculiarities of which


exercised a decisive influence upon the architectural style of

Assos, as well as upon the topographical character of the

inasmuch as

this

site,

material was exclusively employed in the

construction, not only of the most ancient and most impor-

monument, but

tant

Greek town.

of almost

all

the other edifices of the

This stone, the second

in point of

age

among

the three formations of the kind in the Southern Troad, and


the most recent eruption of the volcano of

the product of

Assos, figures

in

Mr.

Diller's geological notes

His subsequent examination

thin sections of the rock,

of

under a microscope, has shown, however, that


rectly

be described as an andesite.

to

as a trachyte.

it is

more

cor-

The groundmass,

which commonly forms but a small portion of the whole,


of a fine granular and

is

porous structure, and of a gray or

occasionally purplish-gray color, the general appearance being

rendered lighter in tone by the presence of innumerable porphyritic crystals of an

which

tals,

give the

opaque or glassy white.

These crys-

at times attain a length of eight or ten millimeters,

stone a superficial

resemblance to granite,

for

which, indeed, the formation at Assos has been taken by


nearly

all

the earlier visitors to the

who have examined


1

site,

and even by those

the reliefs removed from Assos to the

The Geology of Assos, by

J. S.

Diller

Appendix

to the First Report.

INVESTIGATIONS
This resemblance

Louvre.^

AT ASSOS,

1SS3.

increased by a quantity of

is

small crystals of mica and other iron-bearing minerals, the


alteration of

which often produces small

pits

and

stains.

In topographical and architectural respects, the most im-

portant characteristic of this andesite

is

the conformity of

its

cleavage to two distinctly nTarkcd joint planes, the one nearly

Mr. Diller has observed

horizontal, the other nearly vertical.

that the longer axes of the larger crystals,

and

in particular

Hunt, Leake, Richter, Prokesch von Osten, Toujoulat, Texier, and Welcker
works cited in the chapter on the Archseological History of Assos) all
the last mentioned, in a very vivid and humorous
term the stone a granite
1

(in the

trauaccount of the ruins, characterizing the material as " der hassliche


The earliest traveller who has left us
rige, rauhe, graue Granit von Assos."
any account of the site (Manuscript Journal of John Covel, 1677) speaks of the
stone as " a sort of gray marble." Among all those who visited the site previous
.

to the American investigations, the only ones to recognize the all-important volcanic character of the Acropolis, and to designate the stone as a trachyte, are
Webb, Tchihatcheff, and Purearitis. Abbot approaches the truth in calling the

formation a basalt, and

He

is

It

in

describing the walls as built of granite or trachyte.

followed in this respect, as in

may be pardonable

many

that travellers

others,

who

by Schliemann.

could devote but few hours to the

examination of such extensive remains should thus entirely mistake the nature
But what can be said in excuse of the
of the stone of which they are built.
Comte de Clarac {Mitsee de Sadptiire, Antique et Moderne, vol. ii. part ii.,

who, writing as the keeper of the Louvre at the time when the
were removed thither, and describing the sawing asunder of the
blocks by lapidaries under his personal supervision, asserts the stone to be a
granite,
even basing upon this statement an argument in respect to the age
This statement has been accepted by many writers upon Greek
of the temple.
sculpture for half a century, and, notwithstanding the fact that the reliefs have
Paris, 1841),

Assos

reliefs

during this period been exposed to public inspection in the most frequently
Overbeck [Gesckichte der Plastik, 2d edition, Leipzig,

visited capital of Europe,

1869, vol.

p. 98)

says the material of these important

monuments

is

"Gra-

while
nach den Einen, grober aschgrauer Kalkstein nach den Andern "
Liibke, in his History of Sculpture, simply calls it " an ash-gray, coarse-grained

nit

limestone."

Fully to appreciate these errors,

it

must be borne

difference between andesite and trachyte

is

in

mind

that

while the

so slight as to render a definite

determination possible only after microscopical examination, the distinction


the
being based upon the percentage of certain of the constituent minerals

volcanic Assos stone differs fundamentally alike from a primary crystalline rock,
such as granite, and from a metamorphic rock like limestone

archjEOLOcical institute.

52

those of feldspar, are not only approximately parallel to one


another, but are parallel also to these joint planes.

rangement

is,

indeed, sufficiently

marked

This ar-

to suggest a con-

nection between the jointing and the direction in which this


volcanic product flowed at the time of

its

extrusion from the

crater.

The

horizontal plane divides the rock into layers, which at

times closely resemble the stratification of rocks of sedimen-

The

tary origin.

surfaces of the

cliffs,

in the midst of the


vertical cleavage.

owe their existmanner the upright

terraces of the Acropolis

ence to this peculiarity, while in

like

and of the pinnacles of rock which

modern

village,

So variable

is

rise

were determined by the

the resistance of the layers

to the disintegrating action of water, that the surface of the


cliffs is

from

often deeply furrowed, and in building stones quarried

formation a series of parallel depressions

this

is

de-

veloped, resembling those of a weathered sandstone composed


of strata of different degrees of durability.

The

influence of this andesite upon the architecture of an-

cient Assos

is

noticeable chiefly in two ways, determined on

the one hand by the natural cleavage planes, on the other by


the extreme hardness and grittiness of the stone.

The

first

of these peculiarities affected the general design

and the constructive framework.

huge parallelopipedons
ods of wedging.

From

It

was possible

of the material

by the simplest meth-

this fact resulted the

massiveness of

the edifices, the perfection of the city walls,

all

to quarry

due

also

the comparatively early adoption of accurately squared

to

blocks for

their

escarps,

and

hence,

in

particular,

the

frequency of monolithic sarcophagi in the Street of Tombs.


It

appears very probable that the inhabitants of Assos were

the

first

enormous

among
coffers,

the Greeks systematically to employ such


at

once receptacles of the bodies and mon-

INVESTIGATIONS
Liments to the

agus seems

memory

flesh-devouring stone, found


material, as

will

ASSOS,

of the dead.

have received

to

AT

in

1SS3.

Indeed, the sarcoph-

name from

this

53

the famous

the vicinity of Assos

this

be shown in a subsequent chapter, being

used to hasten the decomposition of corpses thus elevated

above the surface of the earth.

Be

this as

it

may, monolithic

sarcophagi were of greater prominence in the cemetery of

Assos than

in that of

The

any other Greek town.

size of the

blocks obtainable permitted the builders of the temple to


the pronaos

meters

with beams and

coffers of stone.

the columns of the Bouleuterion and

There was thus

monolithic.
all

ceil

space having a clear span of nearly three

little

need

In the lower town

Atrium were

Palace

to resort to vaulting

the gates of the town were trabeate, or terminated by the

false

arch

and even subterranean passages, such

as that

the southwest of the Agora, were covered by lintels.

upon

An

in-

exhaustible supply of the stone was to be had just outside the


walls.

The

at least

one hundred thousand cubic meters of

ancient fortifications alone must have contained

The second

peculiarity of the andesite

this material.

was

of influence

chiefly in the carved details of architectural decoration.

been made upon

this

series of experiments

which have

lately

stone by a lapidary, under the supervision of the writer, have

shown

it

to

be one of the most intractable materials ever

chosen for architectural purposes.

While the granular and

porous structure of the groundmass gives to smaller blocks

number and variety


embedded in it make the stone excessively gritty.
To give a homely illustration, it was often remarked that one

a rough and angular fracture, the great


of crystals

day's walking

among

the volcanic rocks of the Southwestern

Troad, or over the ruins of Assos, would wear the soles of

shoes more than a week's excursion

among the limestone forEven with the sharp-

mations on the southern slope of Ida.

ARCH^OLOGICAL INSTITUTE.

54
est tools,

it

a difficult task to cut the andesite to exact

is

emery wheels

surfaces or to delicate forms, and without

impossible to render

The

smooth.

effects of this character-

are evident alike in the choice of the architectural style,

istic

and

it

it is

in the

design and proportion of subordinate features.

would have been altogether out of the question

to adopt Ionic

forms for edifices constructed of such a material, and in no


stance

is

more ornate

this

style

sufficiently

Tread

island of Lesbos and in the Northern

been adopted

From

Assos.

at

It

common on

known

to

in-

the

have

the same cause, the primitive

stone-cutters of this temple found

it

necessary altogether to

omit certain features of the entablature, notably the trunnels


of the regulas and mutules and the terminal plinths of the

and

triglyphs,

members, forms of greater

to adopt, in all small

thickness and less projection than those which, at the time


of building,

were blunt, the

arrises of the shaft

sively flat

the drip of the corona

kyma was much


of the style.

Thus the

had come to be regarded as normal.

larger

tainia

and regulas exces-

was not undercut, and

and simpler than

its

examples

in other

Similar peculiarities are to be observed in

all

the edifices of Assos.

The

resistance

the

andesite

position

in

which

it

is

placed.

depends

weathering

to

upon the stratum from which

greatly

of

of

it

is

quarried, and the

Thus, some of the stones

the lower wall of the great eastern gate,

having been

exposed to the action of percolating water for the

hundred

years,

may

This softening of the andesite


tion of the feldspar
tion.

last five

readily be crumbled with the finger-nail.


is

plainly

due

to the

which enters so largely into

Being unprotected by crystals

hydrated by long exposure

to the

formed into kaolin

On

clay.

its

degenera-

composi-

of quartz, the feldspar is

atmosphere, and thus trans-

the other hand, those carved

stones which have been buried in dry earth, or which, remain-

AT ASS OS,

INVESTIGATIONS

1SS3.

55

ing above ground, have been sheltered from the storms of

dust that

among

the volcanic formations near Assos exercise

the well-known action of the sand-blast, display forms nearly


as sharp and firm as they can have been

In regard to the temple

itself,

umns, and the rings of the


It

many

cut.

capitals, are still perfectly sharp.

may even be doubted whether

from the eastern

when newly

of the arrises of the col-

the outlines of the sphinxes

front, or of the lion

and hind, have been ap-

preciably blunted during the twenty-three centuries or

more

which have elapsed since they received the

from

last strokes

the chisel of the provincial sculptor.

The

following account of the arrangement and of the con-

structive details of the temple supplements

description given in the First Report.^


etition will

be avoided

still,

So

and corrects the

far as possible, rep-

the minute investigations

made

during the two years subsequent to the publication of those


preliminary notes have afforded so

much

further information

concerning the design and execution of this monument, that


it

be necessary to touch upon several points treated in

will

the previous volume.

The

aberration of the magnetic needle, especially noticeable

in the vicinity of the Acropolis,

proved to be so great, that no

process of reversion could warrant entire dependence upon


indications.

The

true pole was hence determined at the

its

mo-

ment when the north star passed the meridian, on the night
The angle thus obtained proved the
of November 28, 1881.
Compare

the reservations which were there


must be prefaced by a reminder that the
time has not come for a thorough and conclusive publication. It is evident
the descriptions of monuments but recently discovered, and in part still hidden
beneath the earth, will be extended, and possibly corrected as the studies
upon the site advance. Indeed, many points are touched upon in this report
only to indicate the direction and scope of the work."
1

First

made

"

Report, pp. 80-121.


following account

The

ARCH^OLOGICAL INSTITUTE.

56

deviation of the main axis of the temple, south of east, to be


15 14' 40".

The

right angles at the corners of the plan were laid out

by the Greek architect with an accuracy surprising even when

examined by modern instruments


of

of precision.

deviation

but six minutes could be detected, the corners of the

northeast and southwest being this

much

too small.

This

deviation in the length of the temple stylobate amounts to

an error of only 54 mm., or about two inches,

one hundred

in the

length of

feet.

The foundations

of the walls

and columns were, without

exception, placed directly upon the native rock, which was not

more than

half a meter beneath the

at its greatest depth, not

At

step.

more than

pavement
1.15

of the naos,

and

m. beneath the lowest

the point last mentioned, namely, the southwest

corner of the building, a massive substructure was provided

by four courses

of

squared stones, measuring respectively

and 25 cm.

35, 30, 25,

in

thickness, each of which projected

about 3 cm. beyond that resting upon


long blocks forming the

first

it.

Elsewhere the

step were placed directly on

This was

the rock, a level bed being cut to receive them.

the case along the greater extent of the northern and western sides

indeed, throughout exactly one half of

the juncture between the

pavement

of the

its

length,

temenos and the

steps of the temple was a juncture between the tooled surface


of the native rock

from

and accurately squared blocks quarried

it.

The

plan of the building, in

exhibits

some

its

present condition (Fig. 4),

points which were omitted from the drawing of

the floor published as Plate 7 in the First Report

notably

the important pry and dog holes, the remains of the concrete foundation of the mosaic extending beyond the cubes

of colored stone, the weathered standpoints of various bases,

J
^

Fig.

4.

Plan of the Temple of

Assos.

Present Condition.

archjEOLogical institute.

58

which probably supported votive statues or


bed

inscriptions, the

lines of the inner course of the cella wall, and, finally,

the fractures of the paving slabs.

The

stones of the steps vary in length from one to some-

what over three meters.


regular,

seldom being

In width the dimensions are more

less

than 0.9 or more than

i.i

m.

In

the case of the stylobate blocks, the upper surface, forming


the pteroma floor, was planned to have a regular

part of

width of

FiG.

5.

I.I

m., and upon the northern side, which was evi-

Stone in Foundations of Temple, with Bed-moulds for Metal


Castings.

dently laid

first,

no deviation from

observed.

The supply

have given

out,

sizes.

Some

this

to

is

be

of accurately quarried stones seems to

is

much

irregularity in the shapes and

of the inner blocks of

fillet,

measurement

however, as the work advanced, and upon the

southern side there

bordering

Isometric.

which proves them

the lower step bear a


to

have been originally

intended for the outside, and to have been rejected on ac-

count of some defect.

upon the

eastern,

Instances of this are the

and the

fifth

stone

fifth

stone upon the western

side,

counting from the north.

most interesting case of the employment

of older

mate-

INVESTIGATIONS AT ASSOS,
rials

in

in the

without doubt, been prepared

59

stones which had,

hewn

default of the regularly

1883.

quarry

presented by

is

one of the inner blocks of the southern lower step, exposed by


the displacement of the superposed stylobate during the exIn the stone in question are cut two bed-moulds

cavations.

sledge-hammers or battle-axes

for the casting of primitive

(Fig. 5).

These moulds are

but they differ in size

20 cm. on the
5

heft, the smaller

cm., in both cases,

shape almost exactly

in

alike,

the larger being 40 cm. long, and

being 32 cm. long.

must have given the

full

The

depth,

thickness of the

there were no ducts by which the molten metal could


object
be run into the hollows, and consequently there can have been
no corresponding upper mould. The process of casting must
have been of the rudest kind. The beds, after having been
;

can have been covered only with a

filled,

flat

stone, so as to

render the upper side of the hammer-heads as even as posand the newly cast implements, after having cooled,
sible
;

must have been


be pried out.
a

block in

so loose in the stone that they could easily

was, of course, impossible to heat so large

It

the

manner

usually

employed by the bronze-

founders of primitive times.

mould

of the

same

kind, for a battle-axe of

Schhemann

much

at Hissarlik

smaller

others
was unearthed by Dr.
are reported to have been found in the lacustrine settlements
of Switzerland,^ and among the prehistoric remains of HunThere are no definite indications
gary,3 and of Sardinia.*
size,

Schliemann (Heinrich),

liios, the

City

and Country of

the Trojans,

London,

18S0, Fig. 601.


2

Gross (Victor), Resultats des Recherches dans

les

Lacs dc

la Suisse occidentale,

Zurich, 1876, Plate 17 ; and a later publication of the same author, Les dernihres
Trouvailles dans les Habitations Lacustres du Lac de Bienne, Porrentruy, 1879,
Plate
"*

I.

Hampel

(Joseph),

Antiquith Prihistoriques de

la

Hongrie,

Esztercom,

1877, Plate 14.


*

Crespi (Vincenzo), // Museo d' Antichith di Cagliari, Cagliari, without date.

ARCHAEOLOGICAL INSTITUTE.

6o

among

age

the

prove

to

mould thus curiously preserved

the

of

the foundation stones of

the temple of Assos

but

the fact that the block, although regarded as unfit for

use at the time of the

original

Persian

was neversome ground

wars,

theless thus ready at the builder's hand, gives


for the belief that

its

had been employed by the primitive

it

Greeks themselves, rather than by any prehistoric race.


be deemed inadmissible,

If

must be assumed
that the squared stone containing the moulds had been
this supposition

found by the Greeks of the

it

century before

fifth

Christ,

while digging for the foundations of the temple.

although the block was

It is interesting to observe, that,

enclosed upon

used as a

all sides,

(not even its outer edge having

the hollows were accurately

step,)

been
with

in

filled

pieces of stone cut to the exact shape of the objects once


cast,

in

the true antique spirit of admitting no imperfect

member in the construction of the sacred edifice.


The stylobate blocks were invariably so tooled
lateral surface of

of the stone, but was restricted

To

the edges.

width from

smoothed

narrow bands adjoining

of the block

plane being

rougher

to

effect this, a slightly depressed

was cut upon the end


this

to

five

bordered

middle

field

with a brush hammer,

by

varying in

fillets,

ten centimeters, which were

an accurate joint was obtained.

until

that the

juncture did not comprise the entire side

tried

and

This emi-

nently rational method of jointing was universally, and in

As we

every age, employed by the architects of Greece.


learn

from the celebrated inscription of Lebadeia,^ which

relates to the
1

stone-cutting and

First published

(Athens, 1S76)

laying of a like

by Koumanoudes (Stephen A.)

more

in the 'Mi\vaiov, vol. iv.

readily accessible in Fabricius (Ernestus),

tura Graeca Commentatiom's Epi^rapkicae, Berolini, 1881


guste), Etudes

Epis^raphiques

details of the stone-work of

siir

l'

pteroma

and

in

De

Architecture Grecque, Paris, 1884.

Greek pavements are

Architec

Choisy (AuAll the

fully described in this

most

INVESTIGATIONS
pavement, joints

thus

ASSOS,

account of a fanciful resemblance

of the central field to a door-opening, framed by

Fig.

6.

6i

1SS3.

were designated by the term

cut

anailiyrosis, evidently on

AT

Employment of Lifting Dog

in laying

its

the lower

lintel

Steps.

Isometric.

and jambs.

Surfaces of contact thus tooled

may be

seen in

the illustrations of the step construction (Figs. 6 and 7), and


in those

showing blocks

corona (Fig.

15).

The

of the inner epistyle (Fig.

last

interesting antique specification,

13)

and

touch was given by grinding the

from the dimensions and forms of the stones,

the tools for cutting them, and the rulers and reddle for testing the accuracy
of their surfaces, to the

methods of casting the metallic cramps and washing

the joints with a solution of nitre.

and 142 of the

inscription.

The verb

6.vaQvpovv occurs in lines 121

ARCH^OLOGICAL INSTITUTE.

62
fillets

with a

sand-saw. 1

flat

The

stone, or

7.

into the joints with a

process of adjustment was the more easy,

as the face of the stone

FiG.

by cutting

which

it

was necessary thus

to re-

Pry-holes and Levers employed in laying the Steps.


Isometric.

move had been


of the middle

greatly decreased in extent

field.

material used at Assos,


1

it

was possible by these means

In the Lebadeia inscription the terms employed


(line 162) and diro^eu (line 125).

TpifxiJ-aroKoyeTj/

by the sinking

Notwithstanding the coarseness of the


to

for these processes are

INVESTIGATIONS AT ASSOS,

6o

1SS3.

abut the stones so closely that not even a needle can to-day-

be inserted between them.

The

blocks of the stylobate were

possible in place,

by powerful

the presence upon

which were cut

lifted,

and

This

derricks.

to receive the

hooks

of the tackling,

The

its

bed

and were
a hori-

in

In exceptional cases, two of these holes appear

zontal plane.

and

proved by

is

the end surfaces of deep, square holes,

so placed that the stone would swing with

upon either

set as nearly as

side.

case was different with the blocks of the lower step,

upon the same

of the foundations

upper surfaces

of these not being

level with them.

The

was

possi-

exposed

to view,

about

it

ble to chisel

upon them an oblique

of the upper

edge of the stone, on that side against which the

next v/as to abut,

arm

of the iron

and

set.

This

and

slot,

in the

middle

through this to disengage the inner

dog by which the adjoining block was


will

be made clear by Fig.

lifted

Where

6.

the

stones of the stylobate, or of the pteroma pavement, are

re-

moved, the greater number of those beneath them are seen

to

have cut upon the upper side one of these slanting notches of
rectangular section, about 4 cm. deep, from

and

10 cm. long.

7 to

is

6 cm. broad,

to

small portion of the lateral joint

surface of the adjoining stone

the lower half of this

is

hereby exposed, and across

seen a sinking of sufficient depth to

receive and firmly hold one of the sharply pointed


lifting

To

arms of the

tongs employed in connection with the derrick tackle.

the notches and sinkings cut upon the stones for this pur-

pose the

name

dog-holes

may be

Although they appear

to

given.

have hitherto escaped the atten-

tion of inquirers into the details of antique architecture, these

marks are
tion

on.

of importance,

inasmuch as they indicate the

direc-

from which the process of laying the courses was carried


In the temple of Assos, for instance,

it

may

thus be seen

ARCHyEOLOGICAL INSTITUTE.

64

masons commenced work upon the northwestern


corner of the building, where the native rock was highest, and
continued from this point, in one direction along the northern

that the

side

The

front, in the other along the

and eastern
courses met

where the

last

western front.

not far from the middle of the southern side,

stone laid

may be

recognized in the eleventh

This

from the southeastern corner.

rately fitted in between those adjoining

last
it,

stone was accu-

and was hence pro-

vided on both ends with sinkings to receive the claws of the


dog-iron, but with no oblique slots for their release.

ther evident, upon the eastern


as the lower step

which served
ones of

the same course

this

It is fur-

front at least, that the

were

laid earlier

blocks

than the inner

being the natural consequence

of the outer stones following an

alignment determined by the

architect.

When

a shifting of the blocks to their final position

necessary, this was effected by

means

was

of a heavy crowbar, the

which was provided by cutting a groove, one or


one and a half centimeters deep, upon the upper surface of the
subjacent stone, at a distance of ten or fifteen centimeters from
purchase

for

the edge of the block to be moved.


erted, the largest stone could

be

slid

By

the leverage thus ex-

along upon the level bed

provided by the course beneath, from the position in which

had been

set

by

joining block.

upon the

it

the derrick to the closest contact with the adIt

cannot be determined, from the marks

whether the crowbar was

stone,

straight,

left

and em-

ployed with a block of wood or metal to transmit the power, or

whether

it

upon the

was curved, so as
stone.

illustration (Fig. 7),

easy

when

to exercise its pressure directly

The former method, shown


would seem

to

at

in the

be the more natural and

but the presence of two grooves in exceptional cases,


the stone

may have had

to

be moved a greater distance,

would, on the other hand, indicate the adoption of a lever of

INVESTIGATIONS
peculiar shape, such as that

AT

shown

side,

and next but one

Fig.

to the

is

to

1SS3.

An

at B.

grooves, cut for a single leverage,

stone in the foundations of the

ASSOS,

65

instance of two

be observed upon a
on the western

cella wall,

See Plan,

southwest corner.

Like the anathyrosis, the grooves cut for the purpose


thus

of

upon

shifting

separate

the

and of every age.

turies

of

and belong

masonry which form the

to

periods widely remote,

anterior to Alexander, and subsequent

tine the Great.

The

stones were

first

noticed by Dorpfeld,^ by

been termed

"

indications of this

Stemmlocher."

aware, they have not been

is

appear

course

They may be observed upon

varieties of rectangular

of Assos,

stones

remains of Greek buildings of every province

the

So

all

city walls

to

may be proposed

method

of prying

whom

they have

far as the present writer

referred

as the

cen-

Constan-

to

by any English

to

writer upon the details of antique architecture, and the

pry-holes

the

technical

name

name

of these

Their great importance to the investigator of Greek

grooves.

be evident from the fact that

remains

will

certain,

from a comparison of the spacing, the position of

it is

possible to as-

blocks once resting upon the course in which such marks


In the temple of Assos, for instance, although the

appear.

stylobate of the eastern and western fronts has entirely dis-

appeared,

posed of

it

six,

may

thus be determined that the one was com-

the other of seven stones, of greater length than

the average of those upon the sides


the blocks which are in like

while the shortness of

manner seen

to

have formed the

lower course of the cella wall indicates this to have begun


with a plinth of considerable height, as was the case with the
cella wall of the
1

Parthenon.

Dorpfeld (Wilhelm), Untersuchungen am Parthenon, in the Mittheiliingen des


vi., Athen, 1881.

deutschen archaologischen Institutes in Athen, vol.

ARCHAEOLOGICAL LYSTITUTE.

66

In the case of both the steps, the joints were bordered up-

on the exposed face by

fillets,

upon the tread

5 cm. in width, and


These projections were cut

averaging

mm.

often rising not less than 5

of the lower step to within about 35

the following riser

mm.

of

and on the stylobate, when not under a

column, they were continued across the entire stone.

purpose was evidently

Their

to obviate, in so far as possible, the

chipping and defacement of the edges during the construc-

And,

tion of the building.

as

was the case with the similar

makeshifts observable upon the pavements of the Propylaia


of

Athens and the great temple of Paestum,

obliteration does not


to the

The

seem

completion of the

to

their subsequent

have been considered necessary

edifice.

blocks of the stylobate, and some of the outer stones

of the lower step, and those of the foundation course be-

neath them, were bonded together by cramps of wrought


iron.

and

These cramps, averaging


light,

21 cm. in length,

seldom exceeding 12 by 16

mm.

were thin
Their

in section.

ends were bent over to more than a right angle, so

when once set, they


stone.
The ends of

that,

could not possibly be loosened from the


the slots cut for their reception were

often curved and pointed, as

shown

in

Fig. 6.

Although

the f]oor of the temple was trodden under foot and exposed
to weathering for wellnigh

two thousand years before being

covered by the earth, the majority of these irons are


position,

and have suffered

little

from

On

rust.

still

in

the other

hand, the lead in which they were set has been in great part

transformed into a white oxide, through the action of the car-

One cramp,

bonic acid of the atmosphere.

of the stylobate,

(No. M. 581).

The

and

is

an exception-

was taken from the southern

ally fine state of preservation,

side

in

now

in

the Boston

Museum

only stylobate blocks not thus bonded

gether upon either end were four

in the

to-

middle of the northern

IXVESTIGATIOXS
As

side.

these were

among

AT

the

ASSOS,

first

1SS3.

be

to

67

laid, it

would ap-

pear that the precaution came to be regarded as more necessary the further the

At

work dvanced.

the corners of the

plan two cramps are attached to each end of the

Such a joining

two stones.

of the steps

by bars

one or

first

of metal

was

of course a disfigurement of the finely tooled surface of the

pavement, and the temple of Assos


exceptional

among

is

in this respect entirely

the constructions of the Greeks.

It is to

be remarked, however, that the dark color of the stone rendered the contrast between the materials less apparent than it

would have been

The

in the case of buildings of

third stone

counting from the west, was cracked


vided upon

in setting,

same shape

as those

upon the upper surface.


repair
22

answer

stylobate,

and was proa cramp

vertical face with

its

of iron of the

..

marble or poros.

upon the southern side of the

its

employed

So well did

this

purpose, that the fissure

would not have been detected


but for the

presence

the

of

iron.

In the case of the


of the lower step
front,

counting from the

the stone
8.

Perforation of the lower Step.


Eastern Front.

line,

or,

correctly,

a circle fFig.

8).

and the tread

at a distance of

block

on the main

north, a hole

Fig.

fifth

is

cut through

in

an oblique

to
in

speak
the

more
arc

of

This hole, opening upon both the riser

20 cm. from the edge, is of


measuring 9 and 12 cm. Its position
corresponded to the side of the column next to the central
intercolumniation.
There can be little doubt that the puroval section, the axes

pose of the eyelet thus carefully chiselled was to provide a

ARCH^OLOGICAL INSTITUTE.

68

hold for the rope by which the animals intended for sacriHoles of the same kind
fice before the fane were tethered.

were cut through the lower stepof the Heraion

at Olympia,^

who have

called atten-

before the columns, and the scholars


tion to this fact give for

The

it

this explanation.

stylobate blocks of the front and of the rear of the

building had

all

been displaced, and, as but two of them could

be found throughout the entire

city,

appear to have been

cut for building purposes, or deeply imbedded in late

Of the two brought

cation walls.

re-

fortifi-

upon the slopes

to light

of the Acropolis, one, measuring 0.93 by 2.245

"")

was from

the east, and by the aid of the shift-holes could be identified


as the third block

by

1.72 m.,

corner.

On

from the north

the other, measuring 1.15

was from the west, and adjoined the southwest


neither were the traces of the columns sufficiently

distinct to indicate the width of the front intercolumniations

more accurately than had been possible by

a calculation based

on analogy.

The most

careful levellings, repeated

slightest trace of an intentional

show the

ture of the horizontals.

were found

to

On

failed to

and regular curva-

the contrary, the steps and floor

be surprisingly even, and the displacement of

the blocks by the

the

and reversed,

many earthquakes which have overthrown


Asia Minor much less than might

cities of this part of

This immunity

have been expected.


fact that

directly

is

to

be ascribed

to the

the pavement of the building rested, in great part,

upon the native

rock,

without the intervention of

deep foundations.
In regard to the construction of the pteroma
laying of the bed of chips beneath
^

Die Atisgrabunoen

zii

Olynipia.

vom Winter tmd Friihjahr 1S77-7S.


und G. Treu.

Berlin, 1S79.

it,

little

floor,

and the

need be added to

Vol. III. Ueherskht der Arbeitennnd Funde

Herausgegeben von E. Curtius, F. Adler,

AT ASS OS,

INVESTIGATIOXS

was

nal intention

it

to

69

On more mature

the account given in the First Report.


sideration, however,

18S3.

appears questionable whether the

cover the stone pavement by a layer of

cement, as was customary in the Doric monuments of

The

irregularity of the jointing in the

temple of As.sos

is

out the building

con-

origi-

pteroma

Sicily.

floor of the

not greater than that noticeable through-

and the

slight differences in level

between

the upper surfaces of the inner blocks and that of the stylobate, as well as the interstices next to the cella wall,

now

so

plainly seen, are, at least in part, to be ascribed to the settling


of those stones

rock,

tive

which did not

immediately upon the na-

rest

upon massive foundations.

or

At Assos,

as

at

Lebadeia, that portion of the lower surface of the pavement


blocks which was above the bed of chips was somewhat more

roughly tooled than the band next

to the outer

contact with the inner stones of the lower step


ference in treatment was

much

less

marked

At Lebadeia

than in the later construction.

in

Assos

this precaution

finger "

little

in

dif-

the archaic

the stones of

the pavement were not allowed to rest upon the


a space " not wider than a

edge and
but this

filling at all,

At

intervening.

was impossible, inasmuch as no

solid

bearing was provided for the inner ends of the blocks, even
in those cases

where they extended across the entire width

of

the pteroma.

The
level

stone

as

pavement

the

sill

naos door

in the interior is 13

in height having,
sill

of the

upper surface

of

the

is

upon exactly the same


stylobate.

cm. above

this,

The mosaic

the difference

without doubt, been equalized by a revetting

of bronze or of marble.

The

pattern of the mosaic has already been described


a
detailed drawing of the corner which remains may, however,
;

be given to show the shape of each small stone (Fig.


1

Lebadeia Inscription,

line 115.

9).

ARCH.'EOLOGICAL INSTITUTE.

70

Fig.

9.

Detail of Mosaic Pavement, Southeastern Corner.

These separate pieces of black and white marble, embedded


were about 5 cm. in depth, and
were originally flushed over with a fine stucco, which comin a thick layer of cement,

AT ASS OS,

INVESTIGATIONS
pletely filled the joints

between them.

1883.

Cubes

71
bright

of a

yellow stone and of a hard-burnt red brick were also found in

Their place

the vicinity.

determined

it

is

in

now be

the composition cannot

only certain that the narrow band which

separated the wave ornament from the

diamond

field of

pat-

tern was of one of these colors.

The arrangement
been made

naos has already

of this flooring of the

to figure in the controversy carried

on by the ad-

vocates of various modes of illuminating the interior of Greek


temples.

has even been held to indicate the form of the

It

imaginary hypaithron, or other opening


mission of daylight.

In the latest contribution to this subject,

remarks concerning the pavement of the temple

Fergusson

of Assos

"The ornamental

the ad-

in the roof for

part of

it is

13 feet wide,

and the

space between the outer face of the cella walls and the pave-

ment

is

nearly as

between 6

may be

and the inner


that there

feet

6 inches and 6 feet 9 inches, or as

the distance between the outside of the walls

lines of the cella at Bassae.

were internal

From

pillars or pilasters,

this

gather

which thickened

the external walls of the cella to the extent of 7 feet at least,

which could only have been done

if it

were wanted

an opaion or some contrivance for lighting the


this

it

must be

replied, that, as

itself

extend so

far

support

was shown upon the plan

the temple given in the First Report (Plate


the mosaic

to

cella."

7),

To
of

the remains of

beyond the ornamental

field,

on

both sides, as to preclude the possibility of pilasters or other


supports having stood where Mr. Fergusson supposes.
over, the layer of cement, in

More-

which the separate pieces

marble were imbedded, remains, as was explicitly stated

of

in the

text (p. 83), to a considerably greater extent than the pattern

shown upon the plan


^

Fergusson (James),

was introduced into Greek

T/ie

reaching quite to the inner side of the


Parthenon

and Roman

an Essay on the Mode by tuhich Light


London, 1883, p. 90.

J'emJ'lcs,

ARCHJEOLOCICAL INSTITUTE.

72

That such supports

enclosing wall.

by

proved beyond a doubt by

also
for

The foundations

them.

Fergusson

as

assume can never have existed

his theory to

obliged

Assos

is

the entire lack of substructures

for all the

placed upon the native rock

is

at

masonry are invariably


was likewise

while, as

staled,

the entire area of the naos was found to be covered beneath


the ancient floor with fine earth, which

bedding

of

is

plainly the original

In other words,

the pavement.

it is

certain that

the ceiling and roof of the naos extended from wall to wall
in a single span,

and that such a clerestory as Fergusson

imagines was impossible.

The suggestion made


floor is of a period later

in the First Report, that the

than the building

itself,

was

firmed by the investigations of the second year.

mosaic

fully con-

At

several

points the filling of earthen chips beneath the cement bedding


of the

mosaic was removed and

sifted,

and the objects taken

therefrom were compared with those which were, in like

man-

found immediately under the stone pavement of the


pteroma. From the latter deposit, which had remained unner,

disturbed since the construction of the temple, were taken a

These were all unglazed,


of rude sherds of pottery.
variety met with in the
rubbed
lustrous,
the
being
of
several
"
On the other hand, the
cities " of Hissarlik.
two oldest

number

majority of the sherds from under the mosaic were glazed.

Among them

were fragments

senting a tragic mask, and

of

moulded

vessel,

repre-

the handle of a delicately painted

vase, evidently of the fourth century before Christ.

piece of one of the original black roofing

tiles

Also a

of the building

itself.

further and most fortunate discovery in this connection

even renders

it

possible to assign an approximate date to the

repaving of the interior.

This was a bronze coin of Gargara,

struck during the

half of

first

the

fourth century before

INVESTIGATIONS AT ASSOS,
Christ,

so

excellently preserved that

mosaic above

it

have been

to

1883.

73

we may suppose

the

one generation,

laid within

at

now preserved in the Museum of Boston (No. A. C. 64). From these


indications the restoration of the temple, which provided new
tiling for the roof as well as a new floor for the naos, seems
most, after the date of

its

This coin

emission.

have taken place at the time of the greatest

to

nence

of Assos,

ruler, the

its

The

when

Aristotle

was

is

emi-

political

living as the guest of

wealthy Hermeias.

foundations of the temple are destined to a speedy

destruction,

its

squared stones being

much

in

the Greek masons of the Southern Troad.

demand among

But the greatest

care was taken by the explorers to remove as


ble of the structure found

as possi-

little

The

in position.

pits

men-

Report were dug only under those

tioned in the previous


parts of the bed of

still

cement where the mosaic

itself

had been

destroyed, and only two of the paving slabs of the pteroma

were

lifted for

the purpose of these examinations.

It

was

as-

certained that the cement had been cast upon a thick layer
of stone chips

and large pieces

the native rock there

of pottery

remained the

still

this

and

which must

in prehistoric ages.

as well as from the interior, several hun-

dred sherds were collected

which tended

between

fine earth

have covered the summit of the Acropolis

From the pteroma,

yet nothing was brought to light

to contradict the opinion

advanced by the writer

in regard to the date of the original construction.

Two

dowel-holes on a block of the pavement immediately

in front of the northern

square plan

without

prominent position.
casting

door-jambs show a narrow stone of

doubt a

The

stele

to

have occupied

this

holes are cut with runs for a lead

these show the dimensions of the base to have been

28 by 55 cm.

detailed description of the foundations of the cella wall

ARCH^OLOCICAL INSTITUTE,

74

The superimposed ma-

has been given in the First Report.

and

sonry was two stones

thick,

for the outer course

was much more

it

is

noticeable that the bed


carefully tooled than

that for the inner, the former surface being thus sunk a few

Pry-holes appear only beneath

millimeters below the latter.

the outer stones, those within not having been accurately


jointed
tions

by means

make

it

of

this expedient.

Both of these indica-

probable that the inside of the wall was cov-

ered with stucco.

On

the southern side the foundations of

the wall are fully 6 cm. below the level of the stylobate.

From
60 cm.,

the width of the stone


it

was at

first

sill

of the

naos door, namely,

wrongly concluded that

this

was also

the thickness of the division between naos and pronaos, and


of the enclosing walls of the sides.

It

has since been ascer-

tained that the stone in question was cut of this width in

order to provide space for the lip of the revetting

sill

placed

upon it, and that the jambs on either side were rebated for
All the walls of the temple were, in fact, of
the same reason.
a uniform thickness of 66 cm., agreeing in this respect with
the antce. At one or two points on the northern side, the
line of juncture

between the inner surface

of the wall

and the

foundation stones could be seen at night-time by the light of


a lamp so held as to send

its

rays in almost the

same plane

with the tooled beds.

The masonry
courses

in

of

the wall itself was, as before said, two

thickness,

the ashlars being consequently only

about 33 cm. in depth. The length of the outer stones of the


lowest course is seen, from the pry-holes upon the foundations,
i.i m., averaging about 84 cm.
must have been demolished at a
comparatively early period. The blocks of which they were
formed, being of a convenient size and squared on all faces,

to

have varied between 0.6 and

The

walls of the temple

provided a most excellent material for the Christian and Mos-

IXrESTICATIOiVS

AT

ASSOS,

1883.

75

lem builders upon the ruins of the ancient town.

So

thor-

oughly had the Acropohs been cleared of these stones, that

specimens were

only four

which

found

They belonged

fied with certainty.

could

measurements of the exposed faces being,


by

by

1.58, 0.81

1.48,

be identi-

to different courses, the

respectively, 0.82

thickness of each was within a few millimeters of

The

The

0.79 by 1.25, and 0.76 by 0.97 m.


33

cm

blocks were thus from three to five times as high as

thick,

and often twice as high as broad.

in all the stones,

Pry-holes appeared

but there were no indications of dowelling of

cramping.
It is

plain that a wall thus

quently bonded by headers

composed must have been

but, notwithstanding this,

never have been really secure, in

The mass

of

masonry,

it

is

this land of

it

fre-

can

earthquakes.

was not weakened by aper-

true,

was but 6.1^ m. high, and was anchored to the entablature upon all sides by the stone beams of the pteroma ceiling.
tures,

On

the other hand, the ratio of

namely, 1:9!

is

its

thickness to

that obtaining in Doric constructions, which,

are in this respect seldom less than


as

9.

its

height

rather below than above the average of

10,

however

large,

and often as much

Moreover, in regard to the length of the wall be-

tween transverse supports, the height being taken as the unit,


we have at Assos a ratio of more than 2|; whereas in the
Doric temples of Greece,

Sicily,

and Magna Grecia we find

the corresponding figures to be not larger than from 2


2h,

to

and the strength of the structure consequently much

greater.

Calculated

use to-day,

we

according to the formulas in practical

find the

stability of

a wall such as that of

the temple of Assos scarcely equal to the requirements of


the case.

renewed scrutiny of the marks of the columns upon the

weathered upper surface of the stylobate furnished some ad-

ARCH^OLOGICAL INSTITUTE.

76

ditional indications in regard

The

last

to the position of the shafts.

block of the upper step upon the southern side, next

to the southeastern corner,

traces of the

now

is

displaced, but

drums which so long stood upon

than the average of the others upon the sides,

The

it.

intercolumniation, thus ascertained, was found

shows

it still

to
its

outer

be wider

clear open-

There was undoubtedly considerable

ing being 1.568 m.

regularity in the spacing of the columns, the width

ir-

the

of

third opening from the southeastern corner not having ex-

ceeded

1.5

m.

Otherwise those intercolumniations which

could be measured with accuracy did not deviate appreciably

from the normal width of 1.532 m,, or 2.447


centre, determined

by

'^-

from centre to

calculation.

The drums unearthed

in the vicinity of the temple,

during

the digging of the second and third years, were of the

same

general proportions as those which had been found before the

They lend

preparation of the First Report.

additional weight

to the conclusions in regard to the height of the shaft,

entire lack of entasis,


lication.

It is

which have been

and

set forth in that

its

pub-

needless to adduce in detail the several hun-

dred measurements upon which these conclusions are based,


as these would merely give a

list

of the accidental and unes-

sential lengths of the separate drums, and of the upper and

lower diameters dependent thereupon.

Suffice

it

to say, that

the diminution was found to average as nearly as possible

mm.

6"]

in the meter, or

15,

this factor

being precisely

the same in the upper as in the lower drums.

The
data,

is

4.78 m.

total

height

column, calculated from these

found to have been within a few centimeters of

The maximum

more than one


the

of the

of this

dimension cannot have been

third of the width of the lower step, while

minimum cannot have been

width of the stylobate.

We

less

than one third of the

thus recognize in the temple

INVESTIGATIONS AT ASSOS,

1SS3.

77

of Assos that ratio which Pliny,^ following Greek traditions,


asserts to have been observed in ancient times

between the

height of the Doric column and the width of the temple plan.

So

present writer

far as the

is

aware, attention has not hith-

erto been called to any instance in which the adoption of this

archaic canon

is

demonstrable.

The lower diameter

the column, with an average of

of

91.5 cm., varied from 90.8 to 92 cm.

the upper, averaging

62.8 cm., from 60 to 63.8 cm.

In so far as regards

its

significance in the history of archi-

tectural development, the exceptional

relation

of the chan-

nelling to the axes of the plan and of the abacus will be

considered, in a subsequent chapter, in connection with the

proto-Doric shaft found in the Necropolis of Assos.

While the beds

of

drums were invariably tooled

the

to

an anathyrosis, the lower surfaces of the capitals were, with


a single exception to be mentioned hereafter, perfectly plane.

This was, without doubt, due to the fact that the joint was
this

in

line

down

to

an almost imperceptible

but was, on the contrary, emphasized by the charac-

teristic

the

case not ground

Doric incision which marked the commencement of

slightly

concave necking.

This single incision, which

increased the opening of the joint to an even width of six or


eight millimeters, was formed by bevelling the edge of the bed
surface of the capital

the slant following the outline of the

channels, and having a width in plan of from 4 to 7 cm.

As
ple

is

regards methods of workmanship, no detail of the temof greater interest than the capital.

whether

The question

as to

most characteristic member of the Doric edifice


was turned upon a lathe, as has been assumed by Botticher,^
this

Pliny, AT. H.,

"

XXXVI.

23 (56).

Das machtige Echinuskyma

des Capitelles,

ist

wohl durch Axendrehung

ARCHAEOLOGICAL INSTITUTE.

yS

was cut entirely by hand,

or

one

is

much

of

importance, in

regard to which the evidence, material and literary, seems not

On

be fully conclusive.

to

centring
capitals

is

visible

upon the bed surface

of every

and the most accurate measurements

made with

cymograph and with

mark

the one hand, the

of

one

some
of the

of the echinos,

strips of lead

bent to the

shape, proved the curve in every case to be absolutely identical

upon

all

sides of the capital

exceedingly

difficult to

a result which

it

would have been

obtain by hand-chiselling.

Moreover, a

passage of Pliny, which seems to have escaped the attention of


writers upon the constructive

the Greeks
lathes

of the sixth

methods of the ancients, shows

century before Christ to have possessed

which were capable

of turning, not only capitals as

dred and

fifty

island of

Lemnos were turned by

of a

a machine of such perfect

tells

advertisement, " a child

modern

upon the neighboring

shafts of the labyrinth

construction, that, as the author

But, on the other hand, there

is

heavy

The one hun-

as those of Assos, but even entire columns.

us, almost in the

could work

it."

words

not a trace upon any one of

the echinoi of such concentric markings as would have resulted from a turning of the stone upon a lathe.

between the greatest projection

of the echinos

The groove

and the under

surface of the abacus even shows lines cut by the chisel in a

contrary direction

so that,

if

the capital

was turned,

at least

was subsequently cut by free-hand.

this quirk of the profile

In view of the difficulty which must have been experienced in


turning so sharply marked a groove against the square of the
auf

dem Bauplatz

Pliny,

TV^. //".,

(labyrinthus).

quarum

in

etc., vol.

XXXVI.

officina

tornarcntur."

Botticher

gearbeitet."

2d edition, Berlin, 1874,

19. 3.

columnis

(Karl),

Die Tektonik dcr Hellenen,

i.

The words

tantum

of the author are:

"

Lemnius,

centum quinquaginta mirabilior

turbines ita librati j^ependerunt,

ut

fuit

puero circumagente

This was probably effected by means of a vertical mandrel.

INVESTIGATIONS AT ASSOS,

1SS3.

79

seem to be an unnatural way


same considerations would
marks which, on some of the capitals,

abacus, this would not, indeed,


of executing the work.

account for the chisel

And

the

are to be seen upon the portion of the echinos immediately

As

adjoining the upper annulet.

upon the bed

surface,

it is

to

mark of a centring
mind that the deter-

for the

be borne

in

mination of this point was quite as necessary in the free-hand


cutting of the necking as in the fitting of the entire block

upon

a lathe.

The sinking

itself

cannot be supposed to have

served for the reception of so gigantic a back-centre as would

fact,

ponderous masses.

to support these

have been necessary

In

emmust reluctantly

stones of this size cannot well be turned without the

ployment

of

chucks upon both ends.

be admitted that the capitals

still

it

temple of Assos do not

of the

While the marks upon the

furnish a proof for either view.

stones

Thus

admit of the assumption that the echinos was

turned from the rough, and the quirk adjoining the abacus,
with the zone contiguous to the annulets, retouched by hand,

they present no decisive indication of the use of the lathe.

Owing
and

to irregularities in the

in the

upper diameter of the

width of the abaci, the echinoi projected

different angles.
in this respect,

and the capitals consequently differed greatly

Yet

it

was found, on graphical com-

parison, that, with a single exception


tioned, the curves of

of the first echinos

presently to be

men-

the echinoi were absolutely identical.

By drawing

the outline

slip of tracing-paper,

and laying

This may be seen from Figure

upon a

over the second, the lines

The

very

Indeed, scarcely two examples were alike

in general appearance.

it

shafts,
at

will

10.

be found to coincide exactly.

third echinos, belonging to a capital of unusual projection,

was lengthened

at its

base by a straight

line,

22

mm.

long

but between the points indicated by asterisks the curve

be found to agree entirely with the

first

two.

Among

all

will

the

ARCHAEOLOGICAL INSTITUTE.

8o

capitals of the temple of Assos, the only

did not conform


in

Figure

to this curve

As may

lo.

is

that

one whose echinos

shown

readily be seen,

it

as the fourth

from

dififers

all

the others, not only in the outline of the echinos, but in the
formation of the annulets, and the extreme shortness of the

necking.

An

explanation of these deviations from the general rule

suggest

will readily

of

modern

itself to

those familiar with the methods

architects and their

cutters intrusted with the

workmen.

making

Each

of the stone-

of the capitals

been furnished by the designer with a templet, in

must have
all

proba-

by which the curves were tested. The


previously hewn upon the blocks of
been
have
must
abaci
stone by the quarrymen, who delivered them to the masons

bility of sheet-metal,

shape of

in the

slabs,

somewhat more than 43 cm.

in thick-

ness, the plan of which averaged 1.193 m. in length and in


These latter dimensions, however, like those of all
breadth.

the details of the temple, which

been tooled by masons


the

different cases;

maximum

1.238 m.

may be supposed

to

have

in the quarry, varied considerably in

minimum observed being 1.18 m.,


Thus the templet, when applied to

the
the

to be inclined from the axis, not only so as to

stone,

had

adapt

the given width of the abacus, but so as to


of the necking, bevelled for the incision, of
base
the

make

itself to

exactly the

of the

same diameter

shaft

for

that of the uppermost

which the capital was intended.

drums, as has been


variation

as

amounting

stated,
to

were themselves subject

not less than 38

mm.

drum
These
to

The angle

which the spring of the echinos formed with the horizontal


large
plane was hence in the first capital shown in Fig. 10 as

Beyond twenty-five
middle
degrees the mason did not venture to go, and when the
upper
the
beyond
cm.
28
than
of the abacus projected more

as

22^

in the

second as small as 15.

Fig. io.

Outlines of Echinos Curves, Anta Capital, and Hawk's-bill


Moulding of Corona.

ARCH^OLOGICAL INSTITUTE.

S2

diameter of the shaft, as in the case of the third capital (Fig.


lo), he was obliged to cut an echinos with an outline considerably longer than that indicated by the templet with which

A share

he was provided.

of this equalization was borne also

by the projection of the necking curve, and by the width of


the former varying from lo to 25 mm., the
the annulets
;

latter
tal

from 50

(shown

to 55

mm.

In the case of the exceptional capi-

as the fourth in Fig.

10),

it

evident that the

is

stone-cutter was without such a templet as that according to

which every other echinos throughout the building was shaped.


The curve was here determined only by the workman's eye.

That

this

was not

particularly accurate can be seen from a

comparison between the actual form, shown by the continuous


line,

too,

and the normal curve, indicated by dots. The annulets,


are of a different design and it is to be remarked that in
;

this capital alone is the

bed surface

tooled, uselessly, to

an

anathyrosis.
It

further resulted from this

method

while the height of the necking

is

one

of

adjustment, that,

of the

dimensions

of the structure, the height of the

exceedingly

namely, from 185 to 216

where the abacus was

felt

to

abacus varies

In one instance,

be altogether too high, and yet

down on account

could not be cut

mm.

most constant

of the impossibility of

shortening the altitude of the column, the vertical faces were

diminished by so tooling the upper surface that the epistyle

upon a scamillus, about 80 cm, square, left


the field, by which they were raised fully

beams

rested only

in the

middle of

25

mm. above

the upper outer edge of the abacus.

crevice between the sides of the capital

bent

lintels

was thought

broad

and the superincum-

to be less offensive,

when seen from

below, than too great a height of the abacus, which could


readily be perceived on account of

The upper

surfaces of

all

its

the abaci

projection.

were bevelled upon

INVESTIGATIONS

AT

ASSOS,

1883.

83

those sides which were placed at right angles to the entablature, in order to obviate a

beams
and a

fall of

8 or 10

an abacus. Figure

chipping of the edges by the heavy

This slant had a width of 6 or 8 cm.,

upon them.

laid

mm.

Compare

the drawing of the top of

11.

f^

-I
/^-

Fig.

On

Upper Surface of an Abacus.

the annulets of the three best-preserved capitals dis-

tinct traces of a

This

II.

c.

tint did

deep vermilion pigment were

to

be seen.

not extend beyond the vertical faces of the an-

nulets, neither to the

necking nor to the echinos.

It

may be

taken as an indication that at least all the smaller mouldings


of the building

point of

were colored.

much importance,

It

proves also, and this

was not thickly primed with stucco, but that the body

ments were applied


stone.

The temple

is

that the surface of the andesite

directly

upon the tooled surfaces

of Assos,

though

built of so

pig-

of the

hard and

ARCH^OLOGICAL INSTITUTE.

$4

gritty a material,

monuments

Attic

is

in this respect to

be compared with the

of the perfected style, rather than with the

archaic structures of the Peloponnesos and of Sicily.


andesite was treated

like

The

the marble of Pentelikos, rather

than like the poros, which its grain more resembles. The
accurate jointing, effected throughout by means of the anathyrosis,

pointed

indeed

to

the

same conclusion

but the

lack of trunnels, and of delicate details in the sculptures of


the epistyle, might otherwise have permitted the assumption
that the minor

members were

supplied,

and the plane surfaces

coated and smoothed with the aid of some plastic composition.


The blocks of the entablature are so weathered that it is not
possible to affirm the
Still,

employment

of

pigments upon them.

the traces upon the capitals suffice to furnish proof that

in the

temple of Assos, as

architectural forms

in all other

Doric monuments, the

were modified and perfected by a poly-

chromatic treatment.

That decorative objects of some light material were affixed


to the columns is evident from the rust marks of iron pins,
once inserted in the groove between echinos and abacus, and
in the joint between the upper surface of the capital and the
epistyle beams. It is not possible to determine whether these
like the shields which
objects were of metal, and fixtures,

once were fastened upon the entablatures of the Parthenon

and the temples of Apollo

at

Delphi and Zeus at Olympia,

or were merely garlands of leaves and flowers with which the

building was adorned on festival days.

Among

the

year relative

most interesting discoveries of


to the temple is a corner of one

capitals (Fig. lo

a).

This

differs

from

the kind, hitherto known, in having


1

to

The most

show

all

other

the

second

of the antae

members

of

the curve of an Ionic, not

careful examination of the capitals of the temple of Assos failed

traces of any painted pattern

upon

the echinos.

INVESTIGATIONS AT ASSOS,
The

a Doric kyma.

fact that

1SS3.

85

was not undercut

it

moulding can scarcely be attributed

to a beak-

to the difficulty of tooling

the stone to this more delicate profile, inasmuch as the Doric

kyma

appears in

corona.

It

upon the upper edge

perfection

full

plain that

is

we have

of the

here to deal, either with a

provincial confusion of the normal details of the two great


styles, or with

either case

a deliberate retention of archaic forms.^

we may suppose the

In

characteristic leaves of the

Doric anta capital to have been painted upon the member.


Viewed at a height of some five meters above the eye, in the
dim and diffused light of the pronaos, this capital must have
been of good effect its well-rounded curves and the in:

clined and projecting

face of

the intelligent care of the

the abacus

shows no traces of dowelling, or

of

bear witness to

The

designer.

small fragment

other metallic attach-

ment.

The
their

epistyle

beams were somewhat

less

bed surfaces than upon their exposed

difference in

workmanship

it

is

finely tooled
soffits.

upon

From

this

possible to determine, with a

certain degree of accuracy, the position of the end joints, as


relative to the axes of the
lintels

column.

It is

from centre to centre of the supports,


respect amounting in

some cases

it

is

to

extend exactly

the deviation

in this

to not less than 15 cm., or

one quarter of the upper diameter of the


hand,

thus found that the

were by no means so planned as

shaft.

On

the other

evident that the triglyphs directly above the col-

That the form of the anta capital was among the last details of the Doric
be definitely established by architectural custom, is indicated by the singularly clumsy and archaic moulding with which it is ornamented in the otherwise fully developed Great Temple of Paestum.
2 This treatment of the face of the abacus of the antas as a slightly
inclined
and projecting surface is a refinement scarcely to have been expected in an
archaic monument.
Although adopted in the Parthenon and Propylaia, and ex1

style to

aggerated in the archaistic temple of Bassai,


or in the temple of Aigina.

it

does not appear in the Theseion

ARCH^OLOGICAL INSTITUTE.

86

umns

same manner

did not vary from the normal axes in the

a fact which will


as did those above the intercolumniations,
frieze.
The height of
the
of
description
be referred to in the
the epistyle

naturally enough, almost constant

is,

would have entailed

ity in this respect

But

difficulties.

in

all

irregular-

constructive

se-rious

other dimensions, and in the form of

every detail, the variations observable in the epistyle, and

throughout the entablature, are so great that it is


impossible to believe that scaled drawings were prepared by

indeed

Nor can

the architect.

the masons have worked according to

an accurately determined system of measurement. The final


shape must evidently have been given to the blocks after
they had been placed in position.
for instance, while averaging 95

100

mm.

So great an

The width

of the tainia,

mm., varies from 85

irregularity in this simple

explicable by the assumption that the total

fillet is

to

only

height of the

members had been


their tops were cut down

stones was altered after the projecting

carved upon them


to a uniform

The

that

is

to say,

level.

epistyle beams, after

having been placed upon the

columns and released from the tackling of the derrick, were


shifted to an exact position and to a close juncture with the
adjoining stone by means of a lever purchasing in pry-holes
To facilitate this process
cut upon the top of the abacus.
the

beam was

slightly uplifted, or rather tilted,

by a crowbar,

the sharp point of which was inserted between the epistyle

and the abacus, in slots, or shift-holes, cut for the purpose


upon the bed surface of the former. These shallow sinkings,
of rectangular profile, generally 3^ by 4I gm. in plan, and from
I

to

\\ cm. deep, are disposed at a distance of from 24 to

50 cm. from the ends of the epistyle beams. They may be


observed, either in Pans or Boston, upon the sculptured
epistyle blocks

removed from Assos.

The same method

of

INVESTIGATIONS

AT

ASSOS,

1883.

g;

shifting will be noticed in connection with


the stone
the coffered ceihng.

Those blocks

beams

of

of the epistyle

which were ornamented with


bordered along the lower edge of their
face by a
narrow fillet corresponding to the tainia upon
the upper edge,
and of about the same dimensions. This
fillet forms an
reliefs are

architectural

framework

for the sculptured composition,


and,

inasmuch as

reliefs

monument.

The unsculpturcd

do not elsewhere appear upon the


epistyle
of Greek buildings, is not to be found
in any other Doric
were

cut,

conformably

such a lower

fillet.

lintels of the

temple of Assos

to the principles of the style,

It is

evident that, in

its

without

ideal form, the

epistyle, like the wall of

which it is the representative, should


have no architectural divisions,
no memberment upon the
face beneath the tainia, or wall-plate.

The

regulas did not have the trapeze


shape assigned to
Texier.i
Their ends were straight and vertical.

them by

As

has been stated

in the First Report,^ the


outer blocks
of the epistyle were provided, along
the upper half of their
inner side, with a rough boss, occupying
nearly one half of
the total height of the beam, and projecting
from
to
5

This peculiar formation

at first

the epistyle beams were three


thenon, a view which was set

20 cm.

led to the supposition that


in

number, as

in

the

Par-

forth in the preliminary de-

scription of the

building.
The investigations of the second
year have, however, given proof of the
contrary.
In the
entablature of the temple of Assos a
constructive system is

now

recognizable which

is

without a parallel in similar fab-

Texier

1
(Charles Felix Marie), Description deVAsie
Mmeure, fait par Ordre
dn Gonvernenunt Fran.ais de 1833 a
1837, et publiee par le Muustere deV
Instruction Pubhque, deux.eme
partie, deuxieme volume.
Paris, 1S49.
The incorrect

statement concerning the shape of the


regula has been repeated
books on Greek architecture.
2 Preliminary Report,
p. 9a

in

many^

text-

88

ARCH^OLOGICAL INSTITUTE.

The epistyle was but two stones


beam occupying nearly two thirds of
rics.

in thickness, the

inner

the width of the

soffit,

yet being less than one half as high as the outer

block resting upon

it

that

is

the inner side of the entablature

Fig. 12.

to say, the

was

at least

The
member of

lintel.

second

20 cm.

less in

Section of the Entablature and Coffered Ceiling of


THE Pteroma.

thickness, and hence

it

was not necessary

to tool

away from

the upper half of the back of the outer epistyle beam, and

from the lower half

of the

back of the triglyph blocks, those

rough and projecting faces which


quarrying.

still

show the marks

Indeed, these bosses, keyed

in,

as

it

of the

were, to the

INVESTIGATIONS

AT

ASSOS,

second course of the inner entablature, are

89

1883.

in the transverse

section seen to have practically formed a broken joint, and

must have considerably increased the resistance which the


mass

of

masonry above the columns could

constantly exercised against

it

earthquakes.

the dislocating effects of

offer to the thrust

by the roofing timbers, and


This

from the section of the entablature. Figure


It

was remarked

cult to

in the First Report, that

will

to

be clear

12.
it

would be

advance any satisfactory explanation of the

triple

diffi-

con-

The saving

effected in the weight

of the facing blocks would have been

more than counterbal-

struction then assumed.

anced by the additional labor required to cut stones, naturally


cleaving to parallel and rectangular planes, into the irregular

shape thus determined


joints

upon the

and the

difficulty of assuring exact

would have been increased through such

soffit

a duplication of the surfaces of contact.

On

the other hand, the

more correct information

relative

to the composition of the entablature gained during the sub-

sequent investigation

is

entirely in

eral character of the design,

the wise

The

economy with which

inner

lintel,

agreement with the gen-

and again permits us

to recognize

the construction was planned.

and the two courses above

it,

were formed

of the parallelopipedons,

most readily obtained

ries of Assos, while the

rough projections upon the back of

in the quar-

the single outer epistyle naturally resulted from the cutting

necessary upon the lower edge alone in order to bring


a straight

and close

sculptured

epistyle

joint

upon the

soffit.

it

to

Although the

block was greatly decreased in weight,

and could consequently be more readily provided and more


easily worked, there

contact,

and a

were

triple

still

the fewest possible surfaces of

memberment

of the entablature

was

thus secured upon the inner side, without the necessity of

introducing for this purpose a low and narrow string-course

ARCHAEOLOGICAL INSTITUTE.

90

behind the corona, and upon the same level with

an expe-

it;

dient which would have presented far greater difficulties in

the coarse andesite of Assos than in the fine-grained and firm

limestones elsewhere employed by Greek builders.

The

true arrangement of the epistyle

beams

is

evident from

the position of the pry-holes on the top of the capital.

and 7

pry-holes of oblique section, a a

upon the outer and inner projections

of

7',

Deep

Fig. 11, were cut

They

the abacus.

provided a purchase for the heavy lever by which the stones,


while hanging from the derrick tackle, were guided to their

ward

From the depth and the broad


sinkings it may be assumed that they

upon the bed.

positions

slant of these

outalso

served to receive the ends of the upper timbers of a staging,

subsequently erected to

The

of the entablature.

was

swung

first

facilitate

epistyle

into position.

from the tackling of the derrick,


tact with the epistyle

by the help
hole

beam,

a, of the outer side,

This stone
it

was

set,

and released

shifted into close con-

above the next column, already


purchasing

of a crowbar,

The corresponding

/3.

work upon the higher parts

inner

in place,

in the transverse pry-

lintel, b,

was similarly

set,

being pried against the outer epistyle by means of a lever


bearing in the slot

7,

and against the adjoining inner

placing the lever in the pry-hole

h.

lintel

In some cases

by

was

it

necessary to shift forward the outer epistyle from the position


in

which

it

was

first

upon the abacus, so

laid

as to bring

it

into the exact alignment determined for the face of the entablature.

This was effected, as in the case of the beam

c,

II, by a leverage exercised from a longitudinal pry-hole,

was

rare,

Fig.

e.

however, that recourse was had to this expedient,

It
it

having almost always been possible to guide the blocks while

hanging from the derrick with


of the outer
to

slots,

a and a

have required no such

sufficient precision

The beam

by means

a, for instance, is

lateral correction,

seen

no pry-hole ap-

INVESTIGATIONS AT ASS OS,


pearing at

The

o).

91

stone to be laid upon the capital was

last

the second inner epistyle,


position, laterally

1SS3.

This was shifted into an exact

d.

by a leverage applied

at

and longitudi-

7',

nally by a purchase provided upon the next capital in that

direction in which the construction


of all the epistyle

beams

to

be

was carried

on.

must have been

The
at

last

one

of

Accurate jointing was assured

the corners of the building.

by cutting an anathyrosis upon

The

laid

all

the surfaces of contact.

indications thus obtained from the pry-holes upon the

top of the abacus were sufficient, not only to prove that the

were two

lintels

of the blocks

in

number, but also

was evident, within certain

of these

to

which formed the inner

determine the width

The height

epistyle.

limits,

from that of the

tooled surface beneath the projecting boss of the outer epi-

The

style.

work

acquisition of these facts permitted, during the

of excavation on the Acropolis subsequent to the

year, and during the remeasurement

whose

original destination

first

squared stones

of all the

had not already become evident, the

recognition of one entire inner epistyle beam, and of six others

more or

less

In height these members, with an

fragmentary.

irregularity of but a few millimeters, averaged 385

mm. Their

width, complementary to that of the outer epistyle blocks, va-

from 495 to 550 mm. The length of the one entire stone,
evidently belonging to the side entablature, was 2.39 m.

ried

This block now serves as one of the jambs


to the

of the

gateway

Turkish fortifications which once occupied the sum-

mit of the Acropolis

it

still

stands upright.

That so few

fragments should remain of the thirty-four inner epistyle

beams
(or

which

formed a

eighty-eight meters,

total length of eighty-four


if

meters

the inner epistyle was cut to a

mitre at the corners, as in the temples of Aigina and Olympia)

is

readily explicable

by the consideration that these

stones, having been without projecting

members, and accu-

ARCH^OLOGICAL INSTITUTE.

92

upon

rately squared

sides,

all

were eminently suited to the

purposes of later builders, Christian and Mohammedan',


looked upon the ancient

The

monuments

who

as a convenient quarry.

blocks of the inner epistyle, like those of the outer,

were somewhat

less

finely

than upon their exposed

upon

tooled

soffits,

their

bed surfaces

and were likewise provided

with shallow rectangular notches, destined to receive the point

''

E3

'^--J

J=X

f^:..
7

^^

q^.:,^

iC

?^/'--

''^

!:
"

V\
Fig.

13.

Jh-':'"''-

Fragments of inner Epistyle Beams, showing Shift-holes


AND Masons' Marks.

crowbar by which they were raised while being shifted

of the

As

was necessary that the widths of the


outer and inner lintels should together make up the total
thickness of the epistyle, namely, 82 cm., they must have
into

position.

been

fitted

it

together upon

the

ground.

Those which had

been matched were occasionally designated by masons' marks.

Three
Figure

of
13.

these

signs,

They

will

the age of the building.

the only ones found, are shown in

be referred to in the discussion of

INVESTIGATIONS
The members
size,

which

AT

ASSOS,

show

of the frieze

18S3.

93

irregularities in point of

exceed those observeti in other parts of the


The triglyphs and metopes found during the

far

structure.

course of the investigations, as well as the spacing of the regu-

upon the blocks of the

las

epistyle, prove the

dimensions

even of those details which were in immediate proximity, and

could easily be compared by the observer

some

in

instances, in the

The width

ten.

of the

that of the largest 575

of seven to

smallest

it

the

four of

have varied,

mm.

ceptional, the nearest to

uncommon,

to

enormous proportion

measuring over 57 cm.


divided into two classes

found was 480,

triglyph

The former dimension was


being 51 cm.

thirty-eight

In

ex-

the latter was not

triglyphs

general, the

recognizable

triglyphs

may be

those of the fronts, averaging 56,


and those of the sides, averaging 52 cm. The corner tri;

glyphs, three of which were found, were of the smaller


a fact which

importance

of

is

in

size,

determining the relative

position of the sculptured epistyle blocks.

In the metopes considerable variations were naturally to be


expected, inasmuch as the equalization of the corners of the
frieze,

and of the front and side intercolumniations, devolved


Still, this fact by no means suffices

mainly upon them.

to explain the great differences in the size of these

members.

The narrowest metope found was

broadest,

cm.

6},

the

mm. The nearest


the maximum 835 mm.

nearly half as large again, namely, 905

approach to the
It is to

the

minimum was

be borne

same

height,

in

mind

680, to

that these blocks were of exactly

and that a variation

could readily be detected by the eye.

remark

is

is

let

us say,

cm.

Particularly worthy of

the fact that adjacent metopes, which can be meas-

ured from

width as

of,

one and

much

as

the

same

epistyle

13 cm., namely,

block,

from 68

to 81

differed

cm.

in

This

proved by the spacing of the regulas on the epistyle sculp-

ARCH^OLOGICAL INSTITUTE.

94
tured

with

horse-legged

centaurs,

Nor was

course of the excavations.


ceptional case.

On

during

discovered

the

this an altogether ex-

the large relief of two bulls, in the Louvre,

the adjacent metopes are seen to have varied

cm.

in width,

namely, from 63 to 6% cm.

Under these conditions it appears probable that the exact


position of the regulas was determined after the members of
This was
the frieze had been set in place above the epistyle.
without doubt effected by leaving the ends

of

the regulas, as

carved while the epistyle blocks were still upon the ground,
somewhat longer than the required dimension, (namely, the

width of the triglyphs to which they appertained,) thus securing the possibility of a subsequent correction.
does, indeed,

seem

to be a difference in tooling

tain of the patches adjoining the regulas,

it

and the

rest of the

although the weathering of the coarse stone

epistyle face,

renders

There

between cer-

difficult to

judge of this point with certainty.^

Truly, the execution of a Doric entablature in primitive

working drawings on a large scale,


computed measurements, was a complicated
work.
The architects Tarchesios, Pythios, and

times, without the aid of

or of accurately

and

difficult

Hermogenes,^ were not without good grounds when they


complained of the irksomeness of laying out the Doric
berment, especially the division of the
details of the entablature

Fortunately,
The

'

we

frieze,

mem-

and the other

thereupon dependent.

are provided, by the difference in tooling

probability that the triglyphs and metopes were placed in position

before the ends of the regulas were cut upon the epistyle tends to disprove
the etymological note of Botticher, Tektonik, p. 204 " Regula ist wohl Uebertragung von Kaviiiv, also Richtscheid oder Norm fiir die Statte der Triglyphen."
Even viewed solely in the light of constructive development this is an altogether
arbitrary assumption. The word Regida, employed by Vitruvius (IV. 3. 4), needs
:

only to be taken in
2

its literal

meaning,

Quoted by Vitruvius, IV.

3.

i.

a straight piece of wood, a


Tarchesios

is

ruler.

probably identical with

Argelios, the builder of the temple of Asklepios at Tralles (Vitr., VII. Pref. 12)

AT ASSOS,

INVESTIGATIONS
on the beds and on the

soffits of

1SS3.

95

the epistyle blocks of the tem-

means of ascertaining that the triglyphs


situated above the columns were but little out of the axes
the maximum error in this respect amounting to less than
ple of Assos, with a

cm.

greater irregularity than this would, indeed, have

been intolerable.

In the jointing of the epistyle

selves, on the other hand, so little attention

beams them-

was paid

correspondence with the axes of the supports, that the

to a
half-

regulas cut upon the ends of the blocks vary in length from

46 cm. the joints themselves must have been so close


that this want of agreement did not force itself upon the
II to

attention

of

the

observer, as

the irregularities in the

did

width of the triglyphs and metopes.

Thus no

aesthetic con-

sideration required an equalization of the lintels, which were

worked, as chanced to be convenient, from the stones provided by the quarrymen.

The

however, to be taken into account

differences in length have,


in the

attempt to determine

the arrangement of the reliefs upon the fronts and sides of


the building.

The backs

of the triglyphs

were so cut that the lower half

formed a rough boss, which corresponded with the projection

upon the upper half of the epistyle beams, and, together with
came into bond with the second course of the main entabThis boss, retaining the marks of the quarrying,
lature.

it,

projects from 8 to 12, and varies in height from 22 to 35 cm.

Compare Figures

12 and

14.

assumed by Schneider and Marini in their editions of Vitruvius,


(Vitr., I. i. 12) and Halikarnassos (Vitr., VII. Pref.
Hermogenes at Magnesia (Vitr.,
12, and Pliny, I/ist. Nat., XXXVI. S- 4. 31)
All three are thus seen to have
III. 2. 6) and Teos (Vitr., VII. Pref. 12).
been Asiatic, and their remarks concerning the Doric system without doubt
as has been

Pythios worked at Priene

express the opinion as to that style prevalent in Asia Minor. This is a point of
much interest in the present connection, for it is to be borne in mind that the

temple of Assos (with exception of the small fane, of much later date, at Pergamon) is the only known Doric temple on the eastern coast of the Aegean.

ARCHAEOLOGICAL INSTITUTE.

96

Both triglyphs and metopes were shifted

to

an exact posi-

applied in pry-holes cut upon the

by means
upper surface of the outer epistyle beams, and show upon
their beds the shift-holes, or rectangular slots which provided
of levers

tion

a hold for the lifting crowbar, the function of which has been
The pry-holes
described in connection with the epistyle.
visible

upon the top

of the epistyle

beams conclusively prove

that triglyphs and metopes were placed in position alternately.

Fig. 14.

The

Triglyph, Face and Side.

triglyphs were not laid

first,

and the metopes then slipped

between them from above, as has been frequently assumed


by writers upon Greek architecture.^ The joints between the
two were hidden by inserting the edges of the metopes into
rabbets, cut upon the sides of the triglyphs in such a manner
in

came
The form

be 6 cm. farther back

that the faces of the former

to

than those of the

of these rabbets

latter.

This time-honored error has been illustrated by a

Expedition Saentifique de Moree, vol.

iii.

plate 10.

steel

Paris, 1831.

which

engraving in the

INVESTIGATIONS
were shaped

AT ASSOS,

1883.

97

the projecting band along the upper

to receive

moulding

part of the metopes, but not the delicate hawk's-bill

terminating them

Figure

The

side view of a triglyph,

in the

general arrangement of the cornice has been described

So great are the inevitable

in the First Report.


ties in

shown

is

14.

the distribution of the mutules, as to

that at least the sofBt of the corona

members

of the frieze

had been

make

irregulariit

probable

was not carved

until the

laid

upon the

epistyle,

and

the position of the individual cornice blocks in relation to

them exactly determined.

The spacing

of

mutules

the

the

lacunaria,

the passage of Vitruvius, before quoted,^

troublesome

could

is

which

in

referred to as so

not well be laid out, or even corrected,

by the stone-cutter, after the blocks of the cornice had been


placed in position.

For

it

was requisite that the length

of

the separate stones which formed the cornice should exactly

correspond with the divisions determined by the irregular

widths of the triglyphs and metopes.

Moreover, the inclina-

tion of the sofBt, forming an acute angle with the vertical face

the entablature, would have cramped the

of

would of

itself

alone have rendered

it

workmen, and

necessary to cut the

deep interstices between the mutules before the blocks were


set in place.

The

surfaces of lateral contact formed

by the anathyrosis

upon the cornice blocks averaged 55 mm. in width; the sinking between them being, in some cases, as deep as three
centimeters.

From

the marks

upon the overthrown stones we may

recognize two distinct methods of lifting these heavy cornice


^

The word

Kiinstler, vol.

3. I, should not, I think, be transby Brunn, Gcschkhte der Gruchischen

lacunaria in this passage, IV.

lated "ceiling," as
ii.

it

usually

is,

(for instance

p. 359, Stuttgart, 1S59,)

mutules.
7

but rather the

soffit

of the corona, or the

ARCHAEOLOGICAL INSTITUTE.

98

means of ropes looped into


broad and deep U-shaped grooves cut upon the lateral joint
the other by means of iron hooks, or
surfaces, a, Figure 15

blocks into position

the one by

dogs, grappling into comparatively small and shallow slots in

the

same

The

position, b. Figure 15.

these, the deep

first of

grooves, are observable in other Doric edifices of early period,

as, for instance,

the temples of Aigina, Paestum, and Se-

Although

linous.

they required

more

stone

much
to

be

cut from the block

than did the

slots of

the second method,

they were far less

When

secure.

projection

the

was not

sufficiently great, or

under-

sufficiently

cut, there was dan-

ger of the loop

ping

oiT

slip-

and by

the swinging of the

l.n

Ends of Cornice Blocks, showing heavy


blocks
the
Attachments of Derrick Tackle A, for
rough
of
edges
the
Looped Rope B, for Iron Dog.
stone must always

Fig.

15.

have sawed upon the

fibre of the rope.

In one instance at

Assos, that of an exceptionally heavy corner piece, the entire

U-shaped boss had broken away, and was replaced by


slot,

into

which the end

of a

beam

second variety, a square or oblong


the most simple form, and

one end

of all

is,

could be inserted.
slot,

a deep

Of

the

about 8 cm. in width,

as a general rule,

is

employed upon

those stones which were lifted by grapples.

Stones upon both ends of which were simple

slots of this

kind

INVESTIGATIONS

AT

ASSOS,

1S83.

99

could not, of course, be laid to a close joint, on account of the

bar of the lifting-iron intervening.

upon one

at least of the

was necessary

two surfaces meeting

to cut,

at a joint, a ver-

channel, through which the grapple could be released and

tical

This channel was made either suff.ciently deep

withdrawn.

hook

in plan for the

when

broad
axis,

the dog

Cornice Block, as tilted in Lifting.


TURNING Grapple.

permit the grapple to be turned

to

and

to

The
somewhat
16.

it

of the iron

slipped backwards, towards the joint

Fig. i6.

as

It

be withdrawn

latter

the

or sufficiently

Release

90*^

in that position, as

arrangement, complicated as

more economical

pass freely

to

on

shown
it

for

its

vertical

in

Figure

appears, was

in respect to stone-cutting,

did not require the horizontal slot to be sunk to so great

a depth as did the former.

One
of the

of the blocks of the cornice, upon the southern side


building,

deserves particular remark.

It

evidently

ARCHAEOLOGICAL INSTITUTE.

lOO

adjoined a stone, inserted between two others already in position,

which

last stone

be considered as the very

to

is

seems

to

This

last laid.

have been swung into the

air,

above the

was noticed that in this case both of


the ends ought to have been provided with vertical channels
through which the lifting-irons could be withdrawn, and that,
entablature, before

as

it

was,

it

it

could not be

The

set.

difficulty

was ingeniously

met, not by lowering the stone and cutting the second channel

upon

it,

but by sinking the release upon the corresponding

joint surface of the adjoining block, here in question.


It

in

appears from the position of

all

the grooves and slots

the ends of the cornice blocks, that these stones were

so balanced as to incline slightly towards their front edge,

which thus touched the bed

first,

and could be adjusted with

great accuracy upon the given line above the triglyphs and

metopes.

Compare Figure

The

i6.

tilt

requisite for

this

expedient was determined by the position of the lifting slots,


which were cut somewhat farther back from the face than the
centre of gravity the exact point being without doubt ascer;

tained by some graphic method based upon the section, as


actual experiment
parenthetically,

was scarcely

that

possible.

It

may be remarked,

modern research constantly tends

to

prove that the unequalled refinements of classic architecture

perhaps the most striking instances


by the corrections

of

which are presented

namely, the curvature

of optical illusions,

of the horizontals, and the inclination and entasis of the col-

umns

were,

like

the solution of static problems such as

those in question, arrived at rather by the means of architectural

drawings on a large scale than by any system of arith-

metical calculation.

dominant

traits of

This was

Greek

in

accordance with one of the

intelligence,

which delighted

in the

expression of an idea by some material representation.

In the variety of methods employed

in lifting the cornice

AT

INVESTIGATIOXS
blocks of the temple of Assos,
of the

wc have another

indication

the structure.

there was as entire a lack of con-

edifice

structive as of artistic unity,

mind

lOI

1SS3.

many independent hands engaged upon

Throughout the
in

ASSOS,

fact

which

to be

is

borne

character and

in the consideration of the provincial

the archaic semblance, yet comparatively recent age, of the


sculptures, so unequal in point of style.

Having been

set in position, the cornice blocks

by iron cramps, sunk

were united

upper surface of the stone at

into the

about the middle of the bed, namely, 70 cm. from the face

These cramps, averaging

of the corona.

were formed of exceedingly tough wrought

by a lead

profile of the hawk's-bill

the cornice

The
fice.

specimen

is

cm.

in

section,

and were

set

Museum

preserved in the

No. M. 578.

at Boston,

The

casting.

iron,

is

shown on a

moulding which terminated

large scale in Figure 10,

b.

cornice block from the southeastern corner of the edi-

Figure

17,

plays almost

all

of particular interest,

inasmuch

as

it

dis-

the marks of dowels, cramps, lifting holes,

etc.,

is

occurring upon the course to which

be considered somewhat
indications which

may

it

in detail, as

belongs.

It

may hence
many

an example of the

be derived even from a single displaced

block.

The
which
its

holes for the grapple irons of the derrick tackle by

heavy stone was

this

ends.

One

of these ends

lifted are to
is

be seen upon both of

exceptional, in being, not a joint

surface, but the exposed face of the eastern corona.

thought preferable
square

method
form

slot

upon

to disfigure this
it,

was

cutting a deep

rather than to depart from the regular

of attaching the tackle

of iron hook,

member by

It

by the adoption of some other

which might, indeed, have been made

to

find a hold

upon the outer edge of the corner mutule, but would

have been

liable to slip unless a

dog-hole had there been cut.

ARCH^OLOGICAL INSTITUTE.

I02

further disadvan-

tage of the plan adopt-

ed was that the hawk's-

moulding of the

bill

front

had

quite

through

be

to

cut

at

a,

Fig. 17, in order to al-

low the grapple, when

hooked
to

B,

into

the

against

flat

lie

slot

the vertical face of the


corona.
to

course,

It is, of

be assumed that the

hole

the

the

in

corona,

face

and

unsightly notch

moulding above

the

in
it,

the

were

and bridged

in

filled

of

over with a stucco of


the same color as the

Upon

stone.

the in-

ner western end of the


block

a joint surface

the

grapple slot

is

provided with a release


channel,

c,

for

turning

the iron, like that previously described, and


illustrated in Fig.

____^
In,,

Fig.

Cornice Block from Southeastern


17.
Upper Surface and End.
Corner.

16.

This proves that the


adjoining stone of the

southern

cornice

was

INVESTIGATIOAS
set

AT ASSOS,
swung

before the corner block was

1883.

103

into position.

On

the other hand, the adjoining block of the eastern cornice,

which was
kind, for

laid afterwards,

we

cannot have had a release of any

see at d, Fig. 17, that a broad channel for this

purpose was cut upon the corner block, without doubt after

had been placed

and the equilibrium

position,

in

it

of the next

stone ascertained by a graphic method, as before explained.

The
ing

it

iron

cramps which attached the block

were sunk

by the

to profit

in carefully cut

stability

to those adjoin-

grooves at e and

which was afforded

In order

f.

to the next stone

of the front by the great depth of bed, from east to west, of


this corner block,

and more particularly

order to anchor

in

the stones of the front cornice together for the purpose of


resisting

the

thrust

lateral

exercised, as

presently be

will

by the inclined course of the pediment


cornice, the cramp at f was placed much farther in from the
face of the cornice than was that at e.
shown,

in this line

the upper surface the bed,

In the tooling of

g,

for

the

stone carved with the gargoyle, serving also as a base for the

corner acroterion, was permitted to project above the

having the

full

tion of the gutter channel to the orifice in the lion's

The

inclination of this surface, g,

of the

may be

end of the cornice block, Fig.

stone, subjected as
ter of the front,

by four

vertical

rest,

slant of the gable, requisite for the continua-

it

was

mouth.

seen in the drawing

The superposed

17.

to the thrust of the terra-cotta gut-

was securely attached


dowels at

h,

h',

h",

to the cornice block

and

h'

On

all

the

cornice stones of the front a plane bed, level with the top
of the hawk's-bill moulding, was cut to receive the blocks

which formed the tympanon


situated in the line

ment

j.

The

veil,
first

the

face of

which was

of the stones of the pedi-

cornice, cut to an acute angle, rested directly

upon the

corner block, having a bearing against an exceedingly stout

ARCHyEOLOGICAL INSTITUTE.

104

dowel, the large hole for which

dowel must have been quite


of the

is

be seen at

to

prevent any slipping

sufficient to

pediment cornice block along

This

k.

its

bed, to which

it

was

exposed through the thrust of the other stones of this course,


At no point in the
lying as they did upon an inclined plane.
entire structure did

more depend upon the expedient

of metal

bondings, and nowhere is greater forethought and care disIn recognition


played in their arrangement and execution.
of the fact that the force tending to outer displacement was

mainly exercised against this dowel,

it

was placed

in a line

by which the corner block was anchored


the cramp f having
to the other stones of the front cornice
for this purpose been removed fully one third farther from
with the cramp

f,

cramp e.
these bondings, bearing upon

the corona edge than was the


relations

and

of

fixed at different times,

we have

In the intimate
different courses

a striking proof of the

thought bestowed by the designer, before the erection of the


building,

upon constructive

details apparently of little signifi-

cance, whose disposition would, in a

be

left

modern work, probably

to haphazard.

From

the sinking of right-angled plan cut in the untooled

boss of stone remaining, at

l,

between the bed

for the pedi-

the wall-plate and

ment cornice and the bed for


it must be assumed that the thrust
was exercised,

in

its

of the

rafters, m,

pediment cornice

lowest block, not directly against the

small surface of the dowel at k, but against the bar of iron

interposed between dowel and stone, and affording a broader

From the size of the


may be judged to have had

surface for the pressure.

the boss

L, this

about 10 cm.

bar
It

sinking in
a width of

was probably not longer than 60 cm., and

did not extend farther to the east than to the west of k, being

imbedded in a socket cut upon the lower outer edge of the


first pediment cornice block, so that the end of the metal bar
was not exposed upon the face of the gable.

IXVESTIGATIONS
The plane bed

AT

ASSOS,

1SS3.

105

for the roof timbering, m, extended along

the sides of the building, upon a level with the upper edge
of

the havvk's-bill moulding, from which

it

distant about

is

Upon this bed lay the wall-plate, n n', into which


53 cm.
were mortised the main rafters, the first of these being situated, as will presently be shown, in the line 00'.
The upper
surface of the cornice, which remained between this plane bed
and the outer edge of the stones, namely, a
meter
that

in width, is

to say,

is

owing

roughly

somewhat

split to

less

about half a

strip

the slope of the roof

tiles

than the 15 slant of the gable,

to the overlapping ends of the tiles.

The

difference in slope will be readily understood

nature of this

by a glance

at

the section of the entablature and roof above the pteroma.

The

Fig. 12.

drawing

Along

line of inclination is

shown by dots upon the

the end of the corner cornice block, Fig.

of

this rough-split surface, at a distance of

the edge,

cut a groove, p

is

p',

averaging

cm.

17.

18 cm. from
in

depth,

the purpose of which was to receive the bent inner edges of


a course of ornamental

tiles,

These

to the cornice blocks

tiles

were attached

subsequently to be described.

of peculiar shape, two of which, q and


cornice block.

The dowels

q',

by iron dowels

appear upon the

of the gargoyle were of square

and that receiving the thrust of

section, those of the antefixes

the pediment cornice were round rods and bars

those of the

lighter terra-cotta course, on the other hand, are of oblong section,

considerably thinner than the ones elsewhere employed,

as they

were not called upon

to resist

any considerable

strain,

but merely to prevent an accidental displacement of the long

and narrow

strips of terra-cotta interposed

between the im-

brices and the stone.

The

antefixes which terminated

the

lines of tegulae

were

each attached to the cornice by two dowels, the corner block

showing borings

for

them

at

r'

and

s'.

The

position of

ARCH^OLOGICAL INSTITUTE.

Io6

these irons furthermore

T and

at

x',

indicates

axes of the

tegulae

and consequently the situation of the

rafters,

the

upon which the imbrices were placed without the intervention of purlins or slots.

shown by dotted
antefixes

The

lines at

are thus

first rafter
o'.

above mentioned

The most

easterly of the

seen to have been removed exactly the

width of one imbrex from the terra-cotta gutter

We

is

of the gable.

must, however, here terminate our examination of this

interesting corner cornice block, reserving the

more

detailed

description of the roofing for a later section.

On

the gable ends of the building the cornice blocks were

smoothly tooled upon their upper surfaces,

in

order to receive

tympanon

the upright slabs which formed the

the stones in question, from the western front,

One

wall.
is

of

so stepped

that the bed thus provided rises to a height of 6 cm. above

the tympanon
that

the

moulding

floor.

From

these indications

depth of the gable


of

exclusive

field,

the corona, was 41 cm.

The

it

is

of

reveal

evident

the

beak

was con-

sequently not greater than the projection of the main cornice,


differing in this respect from the
of Aigina,

where

it

Parthenon and the temple

was necessary

to increase the

width of

the tympanon floor because of the gable groups standing upon

At Assos

it.^

in the

same plane

From
it

is

the wall veil was very nearly (within 2 cm.)


as the face of the entablature.

the traces upon the cornice blocks before mentioned,

further possible to ascertain that the

tympanon

wall

was formed

of stones which varied from 36 to 40 cm. in


During the excavations of the second year three
these stones, belonging to the western gable, were brought

itself

thickness.
of

to light.
^

Of equal width, varying

in this

dimension but a

In the Parthenon and in the temple of Aigina the width of the tympanon

is greater than the projection of the main cornice by respectively one eighth
and two elevenths of the entire thickness of the entablature.

floor

AT ASSOS,

INVESTIGATIONS

1SS3.

\OJ

few millimeters from 92 cm., they were respectively 0.59 and

The

0.837, 0.837 and 1.084, and 1.084 and 1.33 m. in height.

proportions and constructive arrangement of the tympanon

became

wall

angular

veil,

tri-

with a total length of 12.64, had a total height

the

of 1.695 m.,

The

perfectly clear through this discovery.

rise in the three stones

actly 0.268 to the unit of length.

It is

recovered being ex-

furthermore evident,

that a single slab, of pentagonal shape, occupied the centre


of the field,

and that the three stones found were the

ond, third, and fourth upon the north of this.

The

sec-

entire

must consequently have been formed of thirteen stones.


to be a more perfect arrangement than that

wall

This would seem

attained by the adoption of any even

number

of slabs,

must

result, as in the

joint,

with two acute angles at the apex of the wall

Parthenon and Theseion,

stead of the one obtuse angle,

more

easily cut,

which

in a central

and

veil, in-

less liable

to fracture.

In the restoration of the temple, given in the First Report,^


it

was assumed, from a comparison of Doric temples

the same
four.

now

of about

age, that the inclination of the gable was as one in

The

difference between this assumption and the truth

ascertained, namely, between 0.25 and 0.268 in the unit,

amounts
deviation

cm.

to but 11
is

in the total

height of the gable.

nevertheless sufficient to

make

it

This

certain that

the slant of one in four, so easily laid out with entire accuracy,

was not

fixed

upon by the

architect.

Indeed, no simple

arithmetical ratio corresponds with the proportions

ognized.

But on examination

of inclination

is,

it

will

now

rec-

be seen that the angle

with extraordinary accuracy, fifteen degrees,

there being hardly the error of a single minute involved in

the dimensions of these blocks, which themselves represent


nearly half of the entire slope.
1

Preliminary Report, Plate

14.

ARCH^OLOGICAL INSTITUTE.

I08

be remembered that the right angles at the corners

will

It

of the

building were found,

when

tested

precision, to

have deviated but one tenth

true square.

The geometrical accuracy

angle

is

hence not surprising.

by instruments of
degree from the

of a

of the

tympanon

tri-

Its angle of inclination, just

one sixth of that which had been employed

the

in

plan,

could readily be, and doubtless was, laid out by dividing the

We

arc of the quadrant into equal parts.

have here a fur-

shown by Greek

ther instance of the preference so generally


architects for graphic

rather

methods of design,

than arithmetical

for geometrical

In such a gable no

proportions.

eye could have detected the difference between the height


resulting from the ratio of one to four, and that determined

by an angle
significant

of inclination equal to fifteen degrees.

fact,

the

that

designer chose the

It is

rather

latter,

than the former, method of approximation to a purely

aes-

thetic ideal.

The

largest of the three gable blocks (Fig. i8) presents a

curious peculiarity.

and

still

The stone

is

a rejected cornice block,

shows upon the back, which was hidden from view,

the mutules and rectangular steppings of the corona


It

soffit.

appears to have been rendered unserviceable, before the

completion

of the entablature,

end which has been cut

off,

through some fracture of the


or through the mutilation of

None

the projecting hawk's-bill moulding.


details

which

still

show defects

six sevenths of its original length,

account for
faces

its

of those cornice

remain upon the stone, now reduced to

condemnation.

became the bed

One

of the gable

of the
block,

sufficient to

end

joint

sur-

the other being

cut to the required slope, while the back was straightened,

and the moulding removed from the


close jointing

upon the

was sunk into the face

front, so as to

allow of

sides.

of the

former corona, and into the

deep

slot,

75

mm.

square,

INVESTIGATIOyS

AT ASSOS,

109

1883.

opposite side, so as to afford a hold for the lifting tackle.

The two
the

other slabs of the tympanon wall were raised by

same means.

After being released from the grapples, the

stones were shifted into close contact by


applied

in

pry-holes,

L--"

which are

,:
^

means

be observed

to

^'
'

'
.

r-.

';,'

'"

..'

^f

ii,,

..

,.

/
.,jii!-.ii;iii-,.'lii.i^.";!.,;;.,ql!:r',:''

Fig. 18.

of levers

upon the

-iliMi'iH-Vi'l"...;"!'!! ;Jl''-/vli^-,.";."iiii:^

Rejected Cornice Block, recut for Employment in

Tympanon

Veil.

smoothed tops of the cornice blocks

of

the eastern

and

western fronts of the temple.

The tympanon corona was provided with

the customary

hawk's-bill moulding, to separate its soffit from the upright

surface of the

73

mm.

in

tympanon

wall.

It

a projection of 41 cm.

was undercut not

At

less

than

the re-entering angle

ARCH^OLOGICAL INSTITUTE.

no

thus formed the stones were only 15 cm. in thickness,

too

for such a coarse-grained material, although they

little

were

required to support no weight except the light gutter of terra-

Almost
by

cotta.

all

at this point
It is

those which were found had been broken

their

fall.

worthy of especial remark, that the system

of effecting

an accurate jointing by means of leverage was adopted even


in the case

clined bed.

of the gable cornice, which rested

upon an

in-

Pry-holes are visible upon the slanting upper

tympanon

surfaces of the three blocks from the


joints themselves, as in

wall.

Greek gables, were not

all

The

vertical,

but at right angles to the slope.

There were no indications whatever which could

lead to

a belief that the tympanon was ornamented by sculptures.

On

the contrary, the entire lack of such an extensive dowel-

ling as

would have been rendered necessary by the presence

of statues

is

quite sufficient to prove that gable groups never

Upon

existed.

the

upper surfaces of two of the cornice

blocks of the main front, that

to say,

is

upon the

floor of the

eastern tympanon, small pins of metal were found to have

been driven into the stone

without doubt for the purpose

of affixing votive offerings, of no great size or weight.


fortunately,

Un-

these indications do not suffice to convey any

idea of the actual character of such decorations, which

may

have been permanent agalmata, or merely festive garlands


of leaves

The

and

entirely plain.
style

is

flowers.

inner side of the entablature appears to have been

That

this

was the case with the inner

rendered certain by the

lintels

the digging of the second year

were likewise without


the fact

of there

epi-

brought to light during

and that the upper blocks

memberment

is

to

be assumed from

being no fragment of string-courses or

mouldings appertaining

to the structure to

which a position

lyVESTIGATIONS AT ASSOS,
not elsewhere assignable.

is

Ill

188S.

In this respect the temple of

Assos agrees with the Attic Doric of the

fifth

century, rather

than with the archaic style of the farther West.

The

As
its

character of the inner lintel has already been set forth.

to the

second and highest course of the inner entablature,

dimensions may be determined with a certain degree of

accuracy by the

maximum

heights and projections of the

bosses upon the upper half of the outer epistyle and upon the

lower half of the triglyphs, the results thus obtained being

checked by comparison with the corresponding measurements


of the blocks of the inner lintel actually discovered.

Thus

it

becomes evident that the second course was of very nearly the
same height as the outer epistyle (82 cm.), and that the third
or uppermost course was, both in height and average width,

the same as the

first,

or lowest (namely, the inner epistyle),

the variation in no respect having been greater than half an

There can be but

inch.

little

doubt that

this

agreement was

intentional, especially in view of the fact that the total thick-

manner equal to the height


of the epistyle.
It appears probable that some unit of measurement was here embodied. The thickness of the second
ness of the entablature was in like

course did not exceed 35 cm.

The pteroma,

vestibule,

and pronaos of the temple were

covered with a ceiling of coffered stone beams, the recog-

and restoration of which was entirely a work of the


second year. Small fragments of two of these beams were
nition

brought to light during the digging upon the

temple
built

two remaining

in

into the walls of the

site

of

the

their entire length are to be seen

Mosque which stands upon

the

northernmost terrace of the Acropolis,^ and twelve others,

more or

less

lower town

perfect,

were found among the debris of the

namely, seven near the foundations of a portico


1

Frelimmary Report,

p. 93,

Plate 23.

ARCH^OLOGICAL INSTITUTE.

112

hundred meters west of the Greek Bath, and five


modern enclosure about the same distance southeast of

situated one
in a

the Bouleuterion.

Three considerations make

longed to the temple.

it

evident that these stones be-

strong presumption

supplied, in

is

by the presence of several of these beams


among the ruins upon the summit of the Acropolis, where
the

place,

first

the remains of no other antique building are to be found.

Carved blocks known

town among

have been derived from the lower

them a marble

and an inscribed
true,

to

lintel

Greek Bath

capital from the

have,

from a Christian church

it

is

been incorporated into the Mosque which stands upon

the lower terrace of the Acropolis.


of service to the

Stones so shaped as

to

be

Turkish builders were undoubtedly removed

from a considerable distance

to the site of the

Mosque; but

with the fragments of coffered beams before mentioned the


case

is

altogether

refer to the

different.

marble blocks

It

v/ould

be

in the facade of

misleading to

Mosque

the

in

explanation of the remains of a stone ceiling buried in the


earth which covered the plan of the ancient temple, amongst
the ruins of this one edifice, and of no other.

improbable that useless blocks belonging to

It is

more than

any structure

in

the lower town would ever have been carried up this great
height.

On

the contrary, the materials of the temple fur-

nished enough and to spare for the rude mediaeval

fortifica-

tions of the citadel, and, having in part been rolled over the
steep, are

number

met with
of the

in various parts of the enclosure below,

drums, for instance,

the southeast, are half buried


village at the north,

among

lie

upon the slopes

of

the ruins of the Turkish

and were dug out

of the debris

chokes the reservoir beneath the Agora.

Thus

which

the coffered

beams, before mentioned as having been found upon the lower


level,

show, by the very fact of their discovery nearly half a

INVESTIGATIOXS

AT

ASSOS,

II3

1883.

kilometer one from another, that they must have been re-

moved from some common


and plane upon three long
to

Of exceptional length,

centre.

sides,

they were admirably adapted

serve the later builders as jambs and

cline

from the Acropolis, to be used

the Christian edifices

among whose

nothing more unnatural than

to

Nothing

lintels.

down

could be more easy than to drag these blocks

the in-

the construction of

in

ruins they were found

carry them

to

a height

where no building other than the Doric temple ever stood.

The wide

distribution of the coffered

beams

is

fully explained

by the consideration that they must have been the


of the temple to

The evenly squared

fall.

first

blocks

stones of the cella

must have been regarded with covetous eyes by all those


who profited by that edict of Theodosius which authorized
wall

the destruction of heathen temples for the purpose of employ-

ing their materials in the erection of Christian dwellings

and these stones could not be removed

until the ceiling

above

the pronaos and pteroma had been overthrown.

The second point is the character


The marks of hammer and chisel, still
of the coffered

split

in

the stone-dressing.

to

be seen upon some

beams, are precisely the same as those observ-

able throughout the

were

of

temple.

the quarry in

backs of the main

lintels

The untooled upper surfaces


the same manner as were the

the brush-hammered

the sides, forming a shallow anathyrosis,

the blocks of the entablature

beams frequently show near

is

fini.sh

like that

upon

upon

all

and, finally, the beds of the

their ends those peculiar rectan-

gular notches, cut to receive the end of the lifting crowbar

during the process of shifting the stone, which have been


described as existing upon the epistyle blocks and the
glyphs.

These indications are

tri-

sufficient to furnish a definite

proof.

In the third place, there was not in the ancient town any

ARCH^OLOGICAL INSTITUTE.

114

Other building to which a stone ceihng of these dimensions,


or of such a variety of span, could have belonged.

belief

All our

Greek architectural remains leads us to the


that these coffered beams appertained to a temple, and,

knowledge

of

judging from their proportions and style of workmanship, to


an archaic Doric peripteros.

cannot be that a building of

It

such importance as that which

by

attested

is

this

monumental

ceiling should

have so entirely disappeared that no traces

whatever

should have been brought to light during the

of

it

The

course of investigations so thorough as those at Assos.

small size of the town itself scarcely allows us to assume


the existence of two temples of this grandeur within

The

proofs that these

beams belonged

ing of the temple are stated thus

its walls.

to the coffered ceil-

in detail,

because the fact of

the
might otherwise give

the discovery of the stones in such remote localities

majority even in the lower town

rise

to doubts in regard to the correctness of the identification, in


spite of the

agreement

dimension with the plan of

of every

arrangement which the writer believes

to

have ascertained.

As in the Theseion, and other Doric temples, the coffers


must have been supported upon transverse beams extending
from the entablature, above the third course of the inner
to the cella wall,

a distance, including the projection

mouldings upon either


4.06 m. in the vestibule.
recognized.

side, of 2.14

No

m. in the pteroma, and

remains of these supports were

Tooled as they were upon

provided at most with a narrow


it is

side,

of the

kyma

all

four sides, and

along the upper edges,

easy to account for their entire dispersion by later build-

ers, as well as for

the impossibility of identifying

fragmentary remains.

sponding members

From

of other

a comparison with the corre-

Doric

ceilings,

be assumed to have been about two feet


of fact, the

them among

however, they
in width.

may

In point

dimension of 0.615 m., including the projecting

AT ASSOS,

INVESTIGATIONS
mouldings,

such existed,

if

is

found to

fit

115

18S3.

exactly the present

case.

In those beams referable to the pteroma and vestibule the

from 325 to 335 mm. square, averaging 33 cm.,


while the bridges, which separated them, varied from 16 to

coffers varied

Thus

19 cm., averaging a millimeter or so less than 17 cm.

panels

the

were almost exactly half a meter on centres.

The dimension of any set of


more than one centimeter from

three

or

more never varied

the length calculated accord-

This difference, small as

ing to the given mean.

suffices to allow of the slight

adjustment requisite

quite

it is,

in the total

extent of the compartments.

And, on the other hand, the

agreement of the entire plan

to the average

the lacunaria

is

dimensions of

so striking as to place the intention of the

designer, and the restoration of the ceiling, altogether beyond


question.

The width

of the pteroma from entablature to wall

before stated, 2.14 m.

To

is,

as

ascertain the clear span of the

Beam from the Coffered Ceiling of Pteroma.

Fig. 19.

(For Scale, see Fig.

20.)

we have to deduct from this twice the projection of


cyma moulding upon the third course of the inner entab-

ceiling

the

lature

and upon the wall

been about

cm.

Now,

plate.

in a

This dimension must have

compartment four

coffers

length, the four sinkings (together, 1.32 m.), the three

and two
fillet

half bridges (together, 0.68 m.),

(0.035

m.) very nearly

make up

in

whole

and the extra

fifth

the requisite 2.04 m.

ARCHAEOLOGICAL IXSTITUTE.

Il6

The width

of these

compartments may be determined by the

two coffered beams which were found


span from

to fillet

fillet

One

other, 1.52 m.
is

is

From

beam.

22.33 "^O

to

is

rior epistyle,

at

the

the total length of the cella wall (namely,

be deducted the width of the transverse inte-

pteromas from

of the

Like that of the Theseion,

this lintel

to have had the same width as the entab-

The

namely, 82 cm.

21.51 m.,

in

ceiling

the sides and at the rear

which divided the ceilings

may be assumed
ceilings

once determined by the known length

that of the vestibule.

lature,

m.

1.53

beams from the pteroma

shown in Figure 19.


The nutnber of compartments on

of the building
of

of these

Their clear

entire.

was, in one case,

upon the

sides

exactly ten

of

entire length of the

pteroma

the building was, consequently,

times the width of the compartment

plus the main beam.

In like manner, the ceiling of the rear pteroma

have contained
twelve

coffers.

six

compartments of the same

Its actual

length of 12.25 m.,

is

found to

each with

size,

when thus

di-

vided (duly omitting the non-existent sixth beam, and subtracting the

kymas upon

compartments

to

the side entablatures), shows the

have had a mean width of

1.5

The

m.

discrepancy here observable, amounting to less than an inch,


is

entirely negligible in a construction

seen,

which, as has been

everywhere displays much greater irregularities than

this in the

dimensions of individual members.

In regard to the details of the construction,

it

can only be

presumed, from the striking analogy of the Theseion, that a


strong under-tie, higher, but not of greater width, than the
other transverse beams, was carried across the sides,

in

the

line of the rear wall of the cella.

As

in all other

Doric temples, the beams were arranged

entirely without reference to the axes of the supports.

Con-

INVESTIGATIONS
trary as this

is

to

system of design,

our own

statical,

ASSOS,

1883.

and too often mechanical,

transverse beams,

mind, were situated more than

five feet

it

must be borne

the want of agreement with

more apparent

in the

drawn plan

in

above the tops of the

columns, and

been

117

by no means unjustifiable upon aesthetic

it is

The

considerations.

AT

these

(Fig. 22), than

it

is

much

can have

in reality.

The

coffered

beams which formed the

ceiling above the

vestibule were longer than those of the pteroma, each con-

L-.

Beam from the Coffered Ceiling of Vestibule.


This

taining five coffers.

of these stones, found

The

first

2 tn.

Fig. 20.

is

proved by the remains of three

among

and best preserved,

the ruins of the lower town.

shown

in

four entire panels and about half of the

Figure

fifth,

the total length of the block being missing.

20, contains

only
The

37 cm. of

lifting holes,

be referred to hereafter, were cut upon either end

which

will

of the

upper surface

no indication

the one remaining consequently affords

of the original length of the stone.

But that

the coffer beams of this series did actually contain five panels,

ARCH^OLOGICAL INSTITUTE.

Il8
and no more,

is

fully

proved by the second and third speci-

discovered, both of these showing a single lifting hole

mens

cut in the middle of the block, so as to balance


single rope,

and disposed exactly above the centre

it

from a

of the third

sinking.

The length

makes

evident that the ceiling above the vestibule was

it

of these coffer beams, thus determined,

divided into four compartments,

there

consequently being

beams of support extending from the entablature

three

of the

eastern front to the entablature above the pronaos columns.

These main beams, being of so much greater span than those


above the pteroma, would, a priori, be supposed to have had
And in fact, when we suba somewhat greater thickness.
tract

from the

12.25

total

length of the vestibule ceiling, namely,

^hs projecting

n^-'

cymas upon the

side

entablatures

(together about 10 cm.), and the width of four compartments

each containing

we

fii'td

five coffers of

the average size (4

2.535 m.),

beams of support a thickincrease of strength which the

that there remains to the

ness of 67 cm. each,

just that

greater span would seem to require.

It will

be observed that

the division of the ceiling of the rear pteroma into eighteen

and

coffers

five

beams, and of the equally long front pteroma,

or vestibule, into twenty coffers and three beams, permitted


this greater thickness to
latter,

the

being

less

And

it

be assigned to the supports of the

dimensions of two coffers with their bridges

by about 20 cm. than that of three pteroma beams.


may be assumed that the choice of four compartments

in the vestibule, instead of six,

enced by

this consideration,

was

in

some measure

influ-

certainly an extremely rational

and ingenious method of design.

The

division of the vestibule ceiling into an even

of compartments requires a main

beam above

number

the central in-

tercolumniation in the longitudinal axis of the building; an

arrangement which seems

to

have been generally followed

in

INVESTIGATIOXS AT ASSOS,
Greek architecture, appearing

in

1SS3.

19

the temples of Selinous, the

Theseion, the Parthenon, the httle fane of Nike Apteros, and


the temple of Bassai.
of

Assos

much

the

The

vestibule ceiling of the temple

however, from those of

differs,

greater span of

its

coffered

all

these temples in

beams

a peculiarity

which may have been determined by the before mentioned


considerations relative to the thickness of the supports, or

may be

in part referable to a desire

by increasing the

size of the

to gain

and most important section of the

The width

breadth of effect

compartments

in

this largest

ceiling.

of the vestibule ceiling,

from the entablature

of

the front to that above the pronaos columns, was 4.06 m.

This agrees very accurately with the length


ments, which

is

to

be computed from the

of the

compart-

size of eight coffers,

with their seven whole and two half bridges, plus the width
of the extra

by

side,

The

fillet.

making a

Eight coffered beams consequently lay side


total of forty panels in

each compartment.

three lintels which crossed the vestibule must have

been the longest stones employed in the construction of the


temple. That they could be quarried without insuperable difficulty

evident from the existence, in the

is

much

built Bouleuterion, of monolithic shafts of the

which exceeded the clear span requisite


by more than two
in

feet

the lower town,

houses,

is

to

and a

half,

among

be seen a door

less carefully

same

for these

material,

temple beams

being 4.8 m. long.

Elsewhere

the ruins of ordinary dwellinglintel 3.7

m. long.

Nevertheless,

many indications of the care which was taken to diminish as much as possible the weight imposed upon these
supports.
The coffered beams, already relieved of fully one
eighth of their material by the sinkings, were made as thin as

there are

was

at all consistent with strength;

at the

and

in the case of

those

ends of the compartments adjoining the entablatures,

the stone was cut

away from

the ends and outer edge of the

archjEological institute.

I20

upper side so as to form deep

steps.

of support, however, little could

pedient,

which

will

The

Figure 20,

In regard to the beams

be gained by this latter ex-

be readily understood from a reference to

sinkings of the coffers have the boldness pe-

culiar to Doric ceilings of the best period, being cut to a depth

more than

of 17 cm.,

half as

much

as their total width.

Only

those practically acquainted with the details of stone-cutting

can understand

how enormous

a work

is

involved in the exe-

cution of so deep a reveal, in so refractory a material.


great an expenditure of time and labor
tatious

is

in details so

So

unosten-

monuments antedating that


which may be said to have commenced

only to be met with in

debasement of the

style

immediately after the attainment of

To

the age of Iktinos.

illustrate

the shallow coffers of the Prostylos

Bath, or those of the


differ as distinctly

Tomb

its

greatest perfection in

this by Assian examples,


Temple at the west of the

of Apollonios in the Necropolis,

from the lacunaria of the temple, both as

regards design and workmanship, as do the superficial and


pretentious sculptures of the third century from the archaic

works

of the sixth

century before Christ.

in the panels of the

It is

noticeable that

temple the depth of the steps, both upper

made exactly equal to their width, including fillet,


the individual beams being thus conceived as square in section.
Compare the detail. Figure 25. This is the case with
and lower,

is

the small coffers of the pronaos, as well as with those of the

Throughout the

pteroma and vestibule.


coffered beams, that

is

ceiling

the inner

to say those not adjoining the wall or

the entablature, were cut so that each included, at least upon

one of
bridge

its
;

sides, the

fillet

running along the middle of the

the greatest possible width and strength thus being

assured to those stones which were supported only at the


ends.

This

is

evident from

well as of inner beams,

the

all

the fragments of outer, as

former being always without a

INVESTIGATIONS AT ASSOS,

121

ISSS.

while one of the latter in each compartment has two

fillet,

In this

fillets.

were

way

the junctures between the separate

beams

great measure concealed, the bands being slightly

in

sunk, and without doubt painted up to the edges with

some

bright color.

Reference has already been made


coffered

beams of the pronaos

were made

of a

much

to the fact, that in

ceiling the sinkings

the

and bridges

smaller size than in those of the pteroma

and vestibule, thus following a principle of design which

is

evident also in the panels of the Parthenon.

Several frag-

ments

various parts

of these smaller coffers

were met with

being preserved from further

of the town, one, in particular,

Fig. 21.

in

Beam from Coffered Ceiling of Pronaos.


(For Scale, see Fig. 20.)
<

injury

by

its

beam remained
215

mm.

fillets 2

of

all

in its entire length,

in width,

these

the specimens found.

four coffers,

The

Figure 21.

square, were separated by bridges 135

cm.

position in the walls of the Mosque.

mm.

single

sinkings,

broad, with

dimensions being the average

The one

and had a clear span

beam contained
m. This makes it

entire

of 1.42

probable that the pronaos ceiling was divided into three compartments, the two adjoining the antae being each four panels
in width, while the central field
in length.

the

same

Assuming

was square, and seven panels

the transverse

size as those in

beams

the pteroma, or a

have been of

to

trifle

arrangement would very accurately conform

smaller, this

to

the

given

ARCHAEOLOGICAL INSTITUTE.

122
space

the projection of two

cyma mouldings upon

the antae

walls (together lO cm.), the two compartments of four coffers


each, and the central field
(2 X 1-42 m.), two beams of 62 cm.
of seven coffers (2.47 m.)

pronaos, between
possibility

two

that there

is

of four,

making up the

the antas,

total length of the

The

namely, 6.65 m.

only other

may

have been four compartments,

and two of three coffers each

rangement must be regarded

but such an

ar-

extremely improbable, on

as

and awkward character. A square


central field, on the other hand, must have been of good
The design of the pronaos ceiling would thereby be
effect.
brought into connection with that of the vestibule, as the main

account of

beams

its

irregular

former would

of the

almost exactly in the axes of the

lie

The triple division


two middle compartments
from a constructeconomical
would also have been the more
of the latter.

only two transverse supports

ive -point of view,

being

re-

quired.
In regard to the width of the pronaos ceiling, from
the entablature above the columns in antis to the wall above

the door,

is

it

seven coffered

number

plain that this

beams

must have been occupied by


by side

laid side

the dimension of this

of sinkings and bridges agreeing very accurately with

the total of 2.48 m.

The
in

plan of the entire ceiling, as seen from below,

Figure 22

is

shown

while sections of the pteroma and of the vesti-

bule and pronaos are given in Figures 23 and 24, drawn to a

uniform scale for the purpose of comopring the very dissimilar proportions of these

ceiling of the

Figure

12,

For a representation of the


scale, see also the section,

and the isometric elevation. Figure

The arrangement
ble, is of

spaces.

pteroma on a larger

30.

of the coffered ceiling, thus demonstra-

fundamental importance

ground plan

of the temple.

and number

of the

in the consideration of the

It is plain,

not only that the size

compartments must have been determined

Fig. 22.

General Plan of Coffered Ceiling.

124

ARCHuEOLOGICAL INSTITUTE.
upon before a single stone of the
building could be

laid,

the design of the

but that

panels must

have been drawn out, or figured,

by the architect

earlier

than the

The

plan of the stylobate.

dependent ceiling of the


tibule

required

columns

of the

in precisely the

axis as those
to

the

antae

same transverse

columns

of the sides

which they corresponded, be-

beam which

And

was carried above them.


the

fact

f=^

f=\:J=^^.!S'=^.

ric plan,

^]^^^^

to

the

of the

Do-

that, contrary

normal development
Section of Pteroma.

the pteroma of the rear

./?^^^^S"7^"-w

^3

Fig. 24.

and

pronaos to stand

cause of the epistyle

Fig. 23.

in-

ves-

Section of Vestibule and Pronaos.

IXVESTIGATIONS
was made equal

in

AT

ASSOS,

1SS3.

25

width to that of the sides, can only be

explained by the adoption, in these spaces, of compartments

same number

containing the

of

More than

coffers.

this

the arrangement of the ceiling must have been determined at


the

same time as was

tions which
of the
six

is

that simple system of numerical propor-

to be traced

compartments was so calculated

The

cella.

factor

nitude, exceeding two meters, that

mistaken

in

the recognition

centimeters,
this

of

it

of the peripteros

here of such mag-

is

not possible to be

is

The known

multiples.

its

the structure, amounting at most to

irregularities of

in

as to give,

and ten times respectively, the width

and the length of the

The width
when taken

throughout the plan.

have scarcely to be taken

An

regard.

into

some

consideration

agreement of the main dimensions, so

exact as that set forth in the points in question, certainly


furnishes a convincing proof of the correctness of the reconstruction.

In respect to the constructive details of the ceiling, one


peculiarity

requires

still

the proof of the

attention.

While

identification some

fortunately

of the coffered

were shifted by the same lever as the stones

for

beams

of the stylobate

and entablature, and bear the rectangular notches indicative


of that method, others were attached to the derrick rope by
a lewis precisely like that in use to-day.

surface of these latter

is

to

Upon

be seen the narrow

the upper
slot, in sec-

wedge shape, peculiar to this form of


Such a lewis-hole is shown in the plan and top view
tion of inverted

vestibule beam, Figure 20.

the section of these blocks,

From
it

mm.

of the

a careful consideration of

appears that the slots were

so disposed that the centre of gravity should

and 35

tackle.

inside the inclined edge:

this

fall

between 25

proving the chief

iron of the lewis to have been about 6 cm. in width at

narrowest

part,

Figure 25.

As

far as

it

was possible

its

to as-

ARCHAEOLOGICAL INSTITUTE.

126

certain from the fragmentary material under examination, the

beams of the pteroma and pronaos were invariably


by dogs, and only the longer ones, belonging to the
Certain it is, at all events, that
vestibule, by the lewis.
lewis-holes never appear upon those remains of the former,
and dog-holes never upon

coffered
lifted

those remains of the latter

which

could

be

may

identified

Hence

with certainty.

it

be concluded that the

derricks erected

wide span

from

differed

above the

of the vestibule

those

em-

ployed elsewhere in the construction,

shaped

having

wedge-

instead

irons,

of

crampoons, for their tackle,

and being, without doubt,


considerably lighter in
Section of Vestibule Ceiling Beam, showing Lewis Tackle.

Fig. 25.

wood.

When

in the case of

the lifting apparatus which

had been

in

possible,

use for the subthe coffered

But when a new derrick was required by the exigenform was altered

cies

of the vestibule,

The

light stones of the ceiling did not

to

its

to suit the case.

need

to

be attached

the ropes by methods so strong and clumsy as the

shaped grooves, or as the deep

of a lewis

was perfectly

may be assumed
lifted

U-

observable upon the

natural.

that

those five-coffered beams upon

and which were consefrom a single support, belonged to the two

which only one lewis-hole was


quently

slots

Under these conditions the choice

massive cornice blocks.

It

as

the pteroma,

structure was employed also for the laying of

beams.

the

cut,

INVESTIGATIONS
outer compartments

of the

AT ASSOS,

vestibule

12/

1883.

ceiling,

where

the

erection of two derricks, one at either end of the beams,


would have been impossible on account of the insufficiency

the

of

standing

space afforded by

the

side

entablatures.

The beams of the two inner compartments, on the other


hand, must have been lifted and set by the help of two
derricks, two legs of each of which rested upon the entabThe employment

lature above the pronaos columns.

complete sets of tackle, hence to be assumed,

of two

further in-

is

dicated by the fact that the lewis-holes situated at the ends


of the beams, such as that shown in Figure 20, were not cut

common

slanting towards a

centre, but exactly vertical to

the bed surface.

On

the

discovery of the coffered beams

first

ruins of Assos,

it

among

the

was thought that these stones could not

be identified with the temple, inasmuch as the lewis does


not appear to have been employed in any other part of the
This view seemed

structure.

to

find

confirmation

opinion, entertained at the time, that this

so

modern

mode

in

the

of lifting,

appearance, necessarily indicated a later age

in

than that to which the temple can be assigned. The only


lewis-holes which the writer had previously seen among
ancient remains were those in the marble epistyle of the

Olympieion

at

Athens,

temple of Assos.^
unreasonable,

The

is fully

six centuries
first

more recent than the

of these objections, not in itself

met by the above

considerations,

have on this account been set forth at length.

mode
and

of lifting

fifth

was known

to the

And

Greek architects

which

that this

of the sixth

centuries before Christ has been proved, since the

1 The lewis may possibly be that lifting-iron " the teeth of which fit into
holes cut in the stone " mentioned by Vitruvius (X. 2. 2). This, at all events, is
the opinion of Piranesi (Giovanni Battista), Z^ Antichita Romane, (Roma, 1756,)

vol.

iii.

ARCHAEOLOGICAL LWSTITUTE.

128

by the excavaOlympia, where, as the writer can testify from recent


of the investigations at Assos,

commencement
tions at

examination of the remains in the Altis, lewis-holes are to be


Lewis-holes appear
seen on fragments of the oldest ruins.
also upon blocks of the archaic Doric temples a, r, d, and t,
of Selinous.i

named

Selinous, but
offers

j^

of these

particularly

jg

worthy

perhaps the most


erected in the

still

fifth

century before Christ

an example of that very appearance of the lewis-hole

and the U-shaped groove, side by


So
able in the temple of Assos.
invention of the Diadochi,

was

of note that the last

recent of the temples of

in

common

use

it

among

teachers of the Greeks in

side,
far,

which

so remark-

can be shown that the lewis

the Egyptians,^

all

is

indeed from being an

those

earliest

that appertains to the working

of stone.

The arrangement
features,

more or

of the

less

subject

restorations, there

of

its

antefixes,

But as

the stone acroteria.

is

clear in its

main

complete remains having been found

of the terra-cotta bands,

evidently

temple roof

this

tiles,

and gutter, and

part of the building

of

was

and even extensive

to

frequent repairs,

is

uncertainty in regard to some details

construction.

The upper

surfaces of the corona blocks of the sides are

regularly tooled to a slope

somewhat

less

than that of the

roof for a space of about half a meter from the outer edge.

The ends

upon the horizontal bed behind


downward pressure upon the projecting

of the rafters rested

this projection,

all

portion of the corona being thus avoided.

The

imbrices, lying

Hittorff and Zanth, Architecture Antique de la Sicile, 2d ed., (Paris, 1S70,)

Books

3, 4, 5,

and

8,

Plates 16, 44, 47, and 89.

Representations of the lewis appear among the sculptures in the sandstone


quarries of Silsilis. Compare Long (George), Egyptian Antiquities, vol. i. (Lon2

don, 1S32,) or other books upon Egyptian remains.

INVESTIGATIONS
upon the

directly

rafters,

AT

must

ASSOS,

in

1SS3.

129

the lowest course have

overlapped the stone so far as effectually to have prevented

Along the sloping

water from penetrating to the interior.

upper surface of the cornice, at a distance of 18 cm. from


the front, there

is

cut a groove, from two to four centimeters

deep, the purpose of which was evidently to hold the bent

inner edge of a course of ornamental

tiles,

interposed be-

tween the cornice and the terra-cotta antefixes.


fragment of

this original course,

now

in the

Museum

single
at Bos-

ton (P, 4258), was found upon the site of the temple during
the digging of the second year (Fig. 26).
clay,

primed with black,

the lines

ornament.

mm.

thick,

and bears

in relief

meander
During a reswhich

appears to have been

ter the

54

dark gray

of a

toration of the roof,

at least

is

It is of

made

two centuries

af-

54

completion of the

building,

this

moulding

was replaced by a band


of terra-cotta, of about the
26.
Fragment of Tile, with ornamented Edge, from a Course interposed BETWEEN LOWEST IMBRICES
AND Corona.

Fig

same thickness, but

of an

entirely different material,

much more porous and


color.
The peculiarity

lighter in

of this band is that only


edge which were situated immediately beneath the antefixes were ornamented, these sections, exactly

those parts of

its

as long as the original antefixes

pattern, of the usual

were wide, having a wave


Greek type, but quite foreign to the

Doric grammar of ornament.

The

discovery of one of the painted antefixes of the temple,

Figure 27, was mentioned


1

the First Report.^

in

Preliminary Report,

p. 96.

This fine

ARCH^OLOGICAL INSTITUTE.

i.^o

specimen

of archaic

Boston, No. 4149,

Doric terra-cotta, now

in the

Museum

at

remarkable for the richness of the colors,


deep red and black, still to be seen upon it. It is formed of a
is

coarse and porous kernel, coated with a priming of fine clay and

powdered

Fig. 27.

contained

fawn

tint.

in

Antefix.

which gives

The

in

From

slip,

the oxide of iron

a Photograph.

to the surface its delicate reddish

is

local potters in the

and quartz

as a

inner mass contains numerous crystals of the

andesite of Assos, which

by the

known

technically

flint,

thus seen to have been employed

same way

as

was crushed granite

those specimens of terra-cotta from the North-

ern Troad analyzed by Dr. Landerer.^


1

Landerer, in Schliemann's

The

Ilios, p. 21S.

principle of an

INVESTIGATIONS AT ASSOS,

18S3.

131

anathyrosis jointing, so consequentially carried out

stone-work of the structure, appears even

which were moulded with a


of the

bed surface.

all

the

the antefixes,

in

slight projection along the

Compare the

of the

in

edge

In view

section, Figure 28.

tendency of tcrra-cotta to warp in the

baking, this precaution must here have been


of particularly

good

effect,

inasmuch as these

narrow rims could easily be


plane.

Each

an exact

was attached

by two circular dowels

to the cornice

of iron, about

to

filed

of the antefixes

cm.

in diameter, these

evidently having been carried through

the intervening plates of terracotta.

The
pins

holes in which these

were

were

inserted

neatly bored to

.a

depth of

not less than 7 cm.

Their

positions upon the upper

surfaces

of

the

cornice

blocks show the antefixes


to

have been spaced


averaging

distance

at a

be-

tween 61 and 64 cm. from


centre to centre, and thus
Fig

28.

consequently to have been

Antefix Section.

arranged
slightest reference to the

The

the

roofing of Doric temples seems always to have been

constructed without purlins or cross


directly
fixes

without

mutules beneath them.

upon the inclined timbers.

slats,

the

The spacing

tiles

lying

of the ante-

consequently determines also the distance of the rafters

from centre

to centre,

There can be but

little

and the

total

width of the imbrices.

doubt that these

latter

were intended

ARCH^OLOGICAL INSTITUTE.

132
to

be exactly two Greek feet

in width,

dimension commonly employed

The

world. ^

was

set

up

in

in

that

all

is

to say, of a

parts of

the ancient

shown upon the sculptured

flat tile

the market-place of Assos as the

slab

official

which

standard,

has precisely this width of 63I cm. with a length of ']\\ cm.
It is not possible to ascertain the exact age of this interesting
gauge, which will be described in detail hereafter, but
evident that either the slab

temple

itself,

No

when

it

is

as old as the building of the

or that the size of roofing tiles

beginning of the
the period

is

common

at the

century before Christ was retained until

fifth

was sculptured.

this official standard

remains of tegute belonging to the temple were found

in a state of preservation

or their exact shape.

sufficient to

It

is

show

their dimensions,

only certain that they were of

angular section, like that sculptured upon the standard, from

which they cannot have materially

The fragments

differed in width.

of three imbrices belonging to the temple

the only ones referable with certainty to that structure

in the Museum at Boston, P. 4175,


All these are of a coarse-grained terra-cotta,

form part of the collection


4180, 4186.

No two

coated with a lustrous black glaze.


alike.

The

side lips of the first

and third are

section at the juncture with the body, like that

standard, while the

lip of

are precisely

the second

is

of a

curved

shown by the

sharply angular.

The

material of the third also differs from that of the others in

being of a yellow color, and having a tinge of purple in the


The varieties of contemporary manufacture, and of
glaze.
the

tiles

employed

sary, quite suffice to

in

the slight repairs so frequently neces-

account for these differences, which can

scarcely warrant the assumption of so


rations of the roofing.
1

many complete

resto-

As no notches were moulded upon

Dorpfeld, Graber, Borrmann, nnd Siebold, Ueber die Venuendting von Terra-

cotten

am Gcisotmnd Dache griechischer

Bauxtierke.

Berlin, 1S81.

INVESTIGATIONS

AT

ASSOS,

1883.

33

receive the ends of those


the corners of these imbrices to
them, the jointing was rudely
in the courses above or below
of the
cutting off the angles on the inner edge
effected by

upper and outer edge

of the outer lip, so as to

form a

tri-

angular base (Fig. 29).

This clumsy makeshift

upon the spe-

is visible

cimens numbered 4175

and 4180.

The
tem

constructive sys-

of the

temple thus
,

far considered

is

Fig. 29.

Corners of Imbrices, roughly


CUT FOR Jointing.

shown

gable ends of the building were provided with


entrance,
gutters, in order altogether to prevent, above the
upon
objectionable
that dripping which was not considered

in Fig. 30.

The

the time of writing the First Report, the mere


from the garfact of the existence of such a gutter was evident
year. In
first
of
the
goyle brought to light during the digging
the restoration of the temple which is figured in that volume,^
acthis moulding was represented as an anthemion band,
the sides.

At

cording to the analogy of other Doric temples, which show the


crowning member to have been ornamented in the same fashion as coronets, such as that

worn by the Juno Ludovisi.

The

correctness of this restoration was confirmed during the second

year by the discovery of a portion of the original sima, now


in the Museum at Boston, P. 4152. This important fragment,
Fig. 31,

is

formed of a kernel

terra-cotta, coated

which the

details

anthemion are
1 1

mm.

of coarse

and exceedingly hard

upon the face with a


of the ornament
carefully

moulded

of fine clay, in

an archaic astragal and


in

high rehef, projecting

(Compare the section. Fig. 32.)


a dark red glaze are still visible upon the diamond,

from the background.

Traces of

slip

Preliminary Report,

pi. 14.

Fig. 30.

Constructive System of Pteroma.

Isometric.

at asSOS,

IaNVESTIGATIONS
while the shp

itself

of a light red tint.

is

primed with a hard stucco, 2

mm.

The

thickness of the upright

body

is

3 cm., that of the

1883.

The

135
inner side

is

thick, of a yellowish color.

base

something over 4 cm.


The
separate lengths were attached
to the

upper surfaces of the

tympanon corona by

iron pins,

the distance between which in


the single instance capable of

measurement was 53 cm. It


will be remarked that the profile

of the gutter

straight-lined,

of

is

perfectly

having nothing

and graceful

the vigorous

Fragment of Gutter.
From a Photograph.

Fig. 31.

curve which characterizes the

simas of the perfected

style.

The

small

amount

of

water

col-

by these gutters was

lected

dis-

charged through four gargoyles at

One

the corners of the building.

of the fine lion's heads of volcanic


tufa

which performed

was found during the

this function
first

year,

has been fully described and

now preserved

trated.^

It is

Museum

at Boston, S. 1162.

Fragments

and

illus-

in the

of the acroteria

were

found sufficient to convey an understanding of the nature of these promFig. 32.

ter.

Fragment of Gut-

Section and Scale.


1

inent

ornaments, although not to

permit of a complete restoration.

Preliminary Report,

p. 94, pi. 12.

ARCHAEOLOGICAL INSTITUTE.

136

The

ridge acroterion

canic tufa employed

and cut

of 18 cm.,

represented by a block of the same vol-

is

for the gargoyles,

to the

form of a

having a regular thickness

The inner

scroll (Fig. 33).

convolutions are indicated by rectangular incisions, about 8 mm.


broad, which deepen as they retreat from the centre, varying

from a shallow notch to a cut fully 5 cm. deep. The spiral


line thus varies in appearance from a light gray to a perfectly
black shadow. The
perforation

circular

centre of the

in the

correspond-

volute,

ing to the
the

of

tal, is

6cjidaXfx6<;

Ionic

capi-

cut completely

through the

and probably served

Fragment of Ridge Acroterion.

Fig. 33.

stone,

for the insertion

disks of

some more

or gilded metal.

branch

is

thrown

from the

off

a point situated one entire revolution from


tion, the

parallel, incised across the volute.

scroll at

inner termina-

its

juncture being marked by four narrow

upon both

of

brilliant material, such as colored glass

lines,

The treatment

is

nearly

same

the

sides of the slab.

It is plain

that

we have here

to deal with the

upper

scroll

of a central acroterion, closely resembling that of the temple


of Aigina,

now

in the

similar kind have

Glyptothek of Munich.

been found

also

stones of the Parthenon

among

Remains

and, taking these facts in connec-

tion with the representations of archaic temples

gems,

etc.,

there

is

good ground for the

mental anthemion, dcco7ipc

of a

the overthrown

ixox^ a slab of

upon

vases,

monuequal thickness, was


belief that a

regarded as the normal decoration of the apex of Doric gables.

The sky

line of the building

was thus emphasized

at its

most

INVESTIGATIONS

AT

ASSOS,

1S83.

67

by an ornament leaving the forms wliich appear


steles, and frankly treated in

salient point

in the terminations of sacred


profile alone.

In technical respects the fragment closely resembles the


Proto-Ionic capital discovered by the writer upon the site of

The

Neandreia, in the Troad.i

seems

have been employed

to

The

builders of this country.

at

material is the same, and


no other period by the Greek

surfaces are dressed with a fine

brush-hammer of the same kind. In both cases the scrolls


seem to have been laid out by unwinding a cord, to the free
end

of

which was attached a

chisel-point,

in the centre of the scroll as

from a cylinder fixed

an involute,

the

opJuIialmos,

perhaps for this purpose, having been cut entirely through

Above

the stone.

all,

the spiral lines are in both indicated

by peculiar incisions of rectangular section, varying in depth


from a slight sinking to a cutting nearly equal to one third
the thickness of the

stone.
In short, the workmanship is
same school of masonry, and is to be reabout the same period of artistic development,

that of one and the

ferred to

a fact which

will

be

re-

ferred to in the discussion

of the age of the temple.

Of
ria,

fore

to

paw

of

light

of

which

the

attached

to

was the

sphinx or

standing

portion

fragment

single

brought

grififin,

corner acrote-

the

the

the

upon

figure

the

base by

end

was
of

Fig. 34.

the Sima, above the gar1

Fragment OF AcROTERioN.
of Sphinx or Griffin.

Paw

Clarke (Joseph Thacher), A Proto-Tonic Capital from the Site of N'eandreia.


Reprint from the American Journal of Archccology, vol. ii. p. i.

Baltimore, 18S6.

ARCHAEOLOGICAL INSTITUTE.

138

The stone is now in the Museum at Boston,


The carving of the paw displays that mastery in the

goyle (Fig. 34).


S.

1 1

05.

rendering of animal forms which


tures of the temple

and

it is

is

much

so evident in
to

all

the sculp-

be regretted that noth-

ing remains of the body, which would have been especially


interesting as an example of the

Assos

the

in

full

work

the

of

sculptors of

round.

be remarked, that in the choice of subjects for the

It will

corner, as well as for the central acroteria, the temple of

Assos agrees with that


particular reason

But

of Aigina.

the

in

representation

Assos there was

of

the

or the other of these animals having formed

symbol
-

The

sphinx or

connection with the fane of Athena Polias, one

in

griffin

for

the heraldic

of the city.

corner acroterion, like the central

scroll,

the lion's

head, and the before mentioned Proto-Ionic capital from Neandreia, is carved of a fine-grained tufa, obtained
in various parts of the
stratified

by the action of water,

is

of the

same volcanic

mation as the andesite of which the temple

But

is

it

much

softer

and more

easily

from quarries

This stone, though

Southern Troad.

is

worked, and was there-

fore better adapted to the requirements of the earliest


stone-cutters.

So

far as the writer

is

aware, tufa

found among remains of a later date than the


the

century before Christ.

fifth

It

its

Peloponnesos and

is

first

Greek
never
half of

thus bears the same rela-

tion to the archaic architecture of the


to that of the

for-

constructed.

Sicily.

Troad as poros does

As

poros because of

coarseness, so tufa seems to have been discarded by the

masons

of later ages

not wise

for,

on account of

its

friability.

This was

though crumbled by a blow, the resistance

of this stone to the disintegrating effects of the weather


far greater
lion's

than that of the andesite.

head and the delicate

fillets

The forms

is

of the

of the Proto-Ionic capital

IXVESTIGATIOXS AT ASSOS,

1SS3.

139

retain a sharpness unequalled in the sculptures or architectural details executed in other materials.

most careful search was made

for materials

which might

serve to prove the existence of any aperture in the roof of the

building for the purpose of admitting light to the interior.

But nothing was found which could possibly be brought into


connection with such a feature.
All the dimensions of the temple

were remeasured at the

members
were recalculated.
The

close of the excavations, and the averages of those

which show perceptible variations


final results

are given in the following table

Meters.

Length of lower step

30.86

Breadth of lower step

14.58

30.31

Tread of lower step

0.27^

Length of stylobate

Breadth of stylobate

14.03

Exterior of cella, length


"
"
breadth

22.33
7-97

Walls of cella and antae, thickness

Door of

naos, breadth of opening

Interior of naos, length

breadth

0.66
1.65

17.71

6.65

Antse walls, length

3-30

Total width of vestibule, before antse


"
"
pteroma, sides and rear

4-95

Columns on

2.45

"

centres, sides, average

"

front,

Lower diameter of

Upper diameter

shaft,

average

average

of shaft, average

Height of steps, each


"
column, calculated
.

"

shaft, calculated

"

capital,

"

epistyle

average

....

3-03

2.61
0.9

1 1-

0.64
0.28
4.78
4-3

0.48
0.82

ARCH^OLOGICAL nXSTITUTE.

I40

Meters.

Height of

frieze

"

0.78

cornice

0,42

Total height of order,^ including steps, calculated

7.36

Thickness of entablature (epistyle)

0.82

Dimensions of coffered ceiUng, vestibule


"
"
"
sides

4.06

2.14

2.14

2. 48

"

"

"

rear

"

"

"

pronaos

X
X
X
X

12.25
21.51
12.25
6.65

15

Angle of gable slope

The remeasurement

led to the conviction

that

is

it

not

practicable to express the general dimensions of an edifice

constructed of so rough a material as the Assos andesite in


units smaller than half a centimeter.

On comparing

these figures with those given in the First

Report,^ slight corrections will be remarked.

come necessary,

partly through the greater

These have be-

number

of

meas-

urements which have gone to make up the averages, and


partly through the comparison of the steel tape used by the
expedition with an accurate standard,

task kindly under-

taken by Professor William A. Rogers, of Cambridge, Mass.

Attention has already been called to the


the structure the

fact,

dimensions exhibited

that throughout

variations

greater

than those of any other Greek temple with which the writer
is

acquainted.

The above

table in

all

cases states the aver-

age computed from every recognizable block. The labor involved in its preparation may be judged from the fact, that
the case of the columns alone

in

more than one thousand

measurements were taken with rod and


1

tape.

In an anonymous review of the First Report, published in the Americatt


18S2, it was asserted that the height of the order does not

Architect, Boston,

agree with the total obtained by adding together the dimensions of the steps,
column, and entablature. The critic, however, omitted to include the lower
step in his computation.

His

total of

7.oSm. increased by the neglected figure

gives the 7.36 m. of the original table, here repeated.


2

Preliminary Report,

p. 96.

CHAPTER

III.

TEMPLE SCULPTURES.

NOTWITHSTANDING

the fact that during the

first

year of the work at Assos the plan of the temple had

been entirely

and the greater part

bare,

laid

the rude

of

mediaeval fortifications which surrounded the upper citadel

examined, hopes were

might be brought

still

to light

The eleven fragments

entertained that additional reliefs

by further digging

in this vicinity.

which had

of the sculptured epistyle

been already found constituted one of the most valuable


sults

of the undertaking,

and nothing was

from which an extension of

this series

to

be

left

re-

undone

could be expected.

thorough search was consequently made upon

all

parts of

the Acropolis during the early months of the second year, as

has been related in the

The

faces of

all

walls

first

known

chapter of the present volume.


to

have been

built

after the

temple was overthrown were exposed to their lowest courses,


while those masses of masonry which were of sufficient thick-

ness to hide sculptured blocks between scarp and counterscarp were broken up with wedge and hammer.

The

great

square tower adjacent to the mosque was found to contain

no recognizable stones of the temple whatever, while the


sphinxes from the eastern facade proved to be the only work

embedded in those masses of rubble and morwhich protected the uppermost step of the Acropolis

of sculpture
tar

upon the northeast.

But

in

the most recent of the

fortifi-

ARCH^OLOGICAL INSTITUTE.

142

cation walls, hastily piled

namely, those

up of large stones without mortar,

south of the temple

at the

site,

several epistyle

blocks were discovered, one of them sculptured.

block

showing the body of one of the sphinxes from the western


was found, face downwards, upon the sur-

front of the temple

In short, five fur-

face of the earth, outside the citadel gate.

ther fragments of the temple reliefs were discovered during


the work of the second year.
lintel,

Three

of these

form an entire

with a representation of horse-legged centaurs

completes the heraldic sphinxes from the western front


the

last,

light

they are yet of scientific importance, as throwing

upon the significance and arrangement

The

ered by the expedition


of the

The

and

new metope, shows the hind legs of a


Meagre as these results appear in them-

previously known.

one

one

a part of a

galloping centaur.
selves,

number

is

total

new

of the sculptures

count of the fragments discov-

thus brought up to sixteen,

within

Louvre

in 1838.^

of those

removed

to the

pieces in the Louvre represent thirteen separate reliefs

those found by excavation, ten.

The

largest and

most important

of the

newly found sculp-

tures (Fig. 35) represents four centaurs,^ galloping, with upThe design is frankly
lifted fore feet and outstretched arms.

Clarac [Miisee, vol.

ii.

part

ii.)

repeatedly refers to the fragments of sculp-

removed from Assos to Paris as seventeen in number. But if we consider


the relief of the Banquet to be composed of four separate pieces, the total
number of fragments would be eighteen.
2 The archaeological literature upon the subject of centaurs is extensive.
The chief authorities in regard to it are referred to by Colvin (Sidney) in his
tures

Representations of Centaurs in Greek Vase Painting, Journal of Hellenic Studies,


But although he goes as far back as Bochart (Samuel),
i., London, iS8o.

vol.

Hierozoicon, sive de Animalibtis Sacra Scriptura, Londini, 1663,

Caspar), Comnientaires sur

les

Epttres d'Ovide,

La Haye,

and Bachet (Claude


list is far from

17 16, his

complete. The most thorough and learned contributions to the subject in recent
years have certainly been those of Stephani (Ludolf), in the Compte Rendu de la
Commission Imperiale d' Archiologie de St. Petersbourg, 1865 and 1S73.

tit^lMi-

-^411

nVl'ESTIGATIOA'S

AT ASS OS,

1SS3.

43

the differences in position being so slight that

decorative,

the monsters, placed as they were in an architectural frame-

work, and at some height above the eye, must at the

first

glance have appeared almost like repetitions of a convention-

The

alized ornament.
like the heads,

bodies, entirely similar in outline, are,

shown exactly

striving after clearness

in

profile

yet, in a childlike

of representation, the front legs

of

each centaur are placed before, while the hind legs are behind, those of the individuals which they adjoin.

The

pattern-

augmented by this
the
arms are held
exception,
single
With a
overlapping.
the thumbs of the left hand all point upout at length
The right and
wards, the thumbs of the right downwards.
left legs are precisely parallel, being, as it were, shown in
like effect of the

composition

is

greatly

The

perspective.

tails,

great a projection from

made prominent
the

the relief by too

in

background,

fall

in

the

same

curve, nearly to the ground.

The

third

centaur roars, open-mouthed, with a peculiarly

naive and archaic expression.


arms, as

if

in his clenched

The body

of

clean fracture.
excellent,

He

alone has bent one of his

carrying a club or stone

yet nothing

is

grasped

fist.

the

second centaur has been

split off

with

Ot-herwise the preservation of the relief

much

mentary and weathered centaur-blocks which have been

moved

to Paris.

is

superior, for instance, to that of the frag-

Coarse as the stone

is,

the

re-

around

fillets

the heads, the twisted curls of the hair and beards, and the
outstretched fingers, are quite distinct.
in

subsequent

sculptured

lintels,

discussion
that

this

of

the

block

It

will

be shown,

arrangement

was

of

the

probably situated

upon the main facade, above the second intercolumniation


from the southeastern corner, and adjoining that relief
which
discovered by the expedition during the first year

archjEological institute.

144

represents Herakles in combat with the centaurs


fallen

upon him

in the

There can hence be no doubt

ferred to.

ing from the arrows of the hero.^

the position of the heads

they are

flee-

Their precipitous haste

arms flung

well expressed by the

regard to the

in

action in which these four centaurs take part

is

who had

cave of Pholos,^ presently to be re-

into the

and by

air,

the foremost three of which are

stretched forwards in headlong

flight,

while the last

is

turned

1 Some account of this myth was given in the First Report, p. 107
but the
passages of the ancient authors referring to it were not there cited. Those
known to the present writer are as follows. The story of Herakles and Pholos
;

is not mentioned in the Iliad or Odyssey, though evidently referred to in the


pseudo-Homeric Kayui^/os ^ Kepa/xls, 18. There is, however, reason to suppose
that it was included in the narratives of the epic poets and chroniclers of the
notably in those of Peisandros of
seventh and sixth centuries before Christ,
Kameiros, Panyasis, and Herodoros the Pontian. Quintus of Smyrna {Posthom.,
VI. 273, and VII. 107), imitating the manner of Homer, towards the close of the
fourth century of our era, is without doubt following an ancient epic prototype
when he describes the labors of Herakles wrought in relief upon the shield of

among them the combat of the hero with the centaurs of Mount
"when wine and the spirit of strife stirred up these monsters to fight

Eurypylos, and
Pholoe,

house of Pholos." A passage of Stesichoros, preserved by


is the most ancient reference which has been handed down
Among the Attic
It will be quoted in a subsequent passage of the text.
to us.
tragedians this exploit of Herakles is referred to by Sophokles {Tradmt., 1095)
and Euripides {Here. Fur., 181, 364, 1272). We learn from Eustratios (Commentary to Aristotle, Eth. Nicom., III. 5, ed. Camerarius, Francofurti, 1578,

against

him

in the

Athenaios, XI. 499 B,

one of the comedies of Epicharmos was entitled 'Hpa/cAijs b wapa


story is alluded to also by Aristophanes (Frogs, 38, and the scholiast), and told at considerable length by Apollodoros (II. 5. 4), Diodoros (IV.
12. 3-6), and Tzetzes [Chil., V. 111-137), who are our chief authorities for the
Other references are to be found in Theokritos {Idyll.,
details of the exploit.

p. 126) that

^6\cp.

The

VII. 149), Lykophron [Alex., 670, with the commentary of Tzetzes), Ptolemy
{A^ov. Hist., v., ed. Westermann, p. 192), Lucian [Jup. Tragoed., 21), Orpheus
{Argon., 410), Philostratos the

Lemnian {/mag., XVI.), Polyainos

{Strateg., I.

Stephanos of Byzantion (p. 670, ed. Meineke). Further, among the Romans, Virgil {Aen., VIII. 294, with the commentary of Ser\nus, and Georg., 11.
456), Juvenal {Sat., XII. 45), and Lucan {Pharsal., VI. 388, 391).
^ The French authorities attached a much less tragic significance to the two
reliefs of centaurs, belonging to this representation, which were removed from
3. 1),

Assos to the Louvre.

The

clubs and stones with which the devoted combatants

INVESTIGATIONS
backwards, in

profile, to

full

AT

ASSOS,

1883.

45

cast a glance of terror at the

pursuer.
In point of style this block presents a marked contrast to
The most striking difference is
its more archaic neighbor.

while

the centaurs upon the corner block are represented in that primitive combination of man and beast in
which an entire and perfect human being is joined to the
that

trunk and hind legs of a horse, the front legs being human
form of the
in this relief the centaurs show the improved
with human
centaurs
While
monster^ with equine fore legs.
are not unfrequently figured upon archaic vases

fore legs

are
ing
of

armed were held to be the instruments of pastoral music, and the attackcolumn itself but a festive train. The official account of the Director
"Ceci pathe Louvre (Clarac, Musee, vol. ii. part ii.) is delightfully idyllic:

une course de plaisir, et ce que ces centaures tenaient presque tous a la


main gauche et pres de leur bouche, pourrait bien etre une sorte de trompette
course
ou de cornet dont les sons champetres accompagnent et excitant leur

rait

joyeuse."
1 Representations of human-legged centaurs upon archaic vases have been met
Voyage Archeolowith by the writer in the following works Uorow (Wilhelm),
Witte (Jean
I. 6, and IV. 2.
PL
Paris,
Etrurie,
1829,
I'Aiicienne
dans
gique
:

Joseph Antoine Marie de), Pelee et


engraved in the Montimenti Liediti

Thetis, Annali,

for

1832, vol.

Roma, 1832, pp. 91-127,


Roma, 1829-32, pi. 37.

i.,

Micali (Giuseppe), Storia degli Antichi Popoli Italiani, Firenze, 1832, pis. 19,
Inghirami (Francesco), Etrusco Museo Chiusino, Firenze, 1833-34,
20, 95.
Maximis (Franciscus Xaverius de), Musei Etnisci quod Gregorhts XVI.,
pi. 84.
in Aedibus Vaticanis Constituit

(Joseph

Emmanuel

Bntxelles, Bulletin, vol.

ix.,

Roulez
Academic de
Campana (Giovanni Pietro),

Monimenta, Romae, 1842,

QlVx^Xz:^),- L' Education

2me

vol.

d'Achille, pi.

partie. 1842.

ii.

i.

pi. lOO.

E.

Antiche Opere in Plastica, Roma, 1842-52, part 2, pi. 22. Micali (Giuseppe),
Momimenti Inediti, Firenze, 1844, pi. 27. 4. Michaelis (Adolph), At/ieuische
Vasen, Archiiologischer Anzeiger,
1861,

No.

14.

Gamurrini (G.

No.
F.),

149, 150; Archdologische Zeitung, Berlin,

Un

Antico Sepolcreto in Arezzo, Annali,

entitled Scavi d' Arezzo, by the same


Roma, 1S69, p. 72. Heydemann (Heinrich Gustav
Dieudonne), Vasensammluug des Museums zu Palermo, Archdologische Zeitung,

Roma,

1872, p. 279;

compare the notes

writer, in the Bullettino,

Salzmann (Auguste), Necropole de Camiros, Paris, 1875, P's- 26,


Colvin, Centaurs in Greek Vase Painting, quoted above, pis. i and 2,
Puchstein (Otto), Kyre7iaische Vasen, Archdologische Zeitung, Berlin,

Berlin, 1871.
27, 39.
fig.

4.

1881.

Special search

among

the catalogues of vase collections would with-

10

ARCHAEOLOGICAL INSTITUTE.

146
and gems,^

Assos

in

sculpture

they are exceedingly rare.

The

Herakles and Pholos, discovered by the expedition, is the only known example of the occurrence of such
lorms in any work of considerable size, or of monumental
relief of

character.
of

figurine,

found by Ross upon the AcropoHs

Athens, and a small bronze

out doubt reveal

many

unearthed at

lately

relief,

Museo Etrusco, Rome, is an inedited


one of the human-legged centaurs of
which is shown in Fig. 36. Besides two fragments from Kameiros, published by
Salzmann in the work quoted above, and the vase given by Colvin, pi. 2, there
others.

vase with representations

in

In the

low

are in the collection of the British

relief,

Museum two

fine inedited

vases with repre-

sentations of human-legged centaurs.

These are numbered B. 116, and B. 420.


The scene illustrated by the latter is the reception of
Herakles by Pholos. The writer has observed unpublished vases showing human-legged centaurs in
the Louvre and the Museum of Berlin
but as the
purpose of the present note is merely to provide

?-l'

proof for the statement

meration

is

the text, a further enu-

in

unnecessary.

A curious uncertainty of form a hesitation between human and equine members


is noticeable
in

Human-legged

FiG. 36.
a

Vase

in

the

S'''"-

J^"^

Museo Etrusco,

centaurs with

vase published by
Vasi Corintii,

same combination appears upon one

H el big

human

gems

fore legs termi-

shown upon an archaic


Wolfgang), Itnitazioni di

Roma,

Anuali,

of the

referable to a period of tran-

nating in horse's hoofs are

Centaur.
Upon

some representations

in the

1863,
British

i.
The
Museum, re-

tav.

ferred to in the following -note.


1 In the British Museum are two gems representing human-legged centaurs
one from the Hamilton, the other from the Castellani collection. At the time of
Woodcuts of them are
writing they are not designated by catalogue numbers.

given by Colvin

in the

Essay before quoted,

figs.

and

3.

Two

other

gems

are

by Micali, Storia, pi. 46. A striking peculiarity of these latter is


the one, with the front legs
that both of the human-legged centaurs are winged
terminating in talons, having the wings extended from the human shoulders, the
other from the horse's back, Pegasos-like. These additions open a wide vista of
monstrous formations, and prove the agglutinative character of such types, which
may very probably have arisen, as Mr. Murray has suggested to me, through the
illustrated

combination of various heraldic symbols, like the quarterings of our coats of


arms. A fifth gem, showing a human-legged centaur, and published by Gori
(Antonio Francesco), Museum Florentinum, Florentiae, 1731-66, vol. ii. pi. 39,
appears, as well as can be judged from the exceedingly mannered engraving, to

AT

INVESTIGATIONS
Olympia, have been mentioned

ASSOS,

in the

1SS3.

Preliminary Report

47

as

being the only parallels to this representation which are to be


found in the wide field of Greek decorative sculpture. To

may be added two Etruscan

this

bronzes,- between three

and

tcrra-cotta figurine

inches in height, and, notably, a


now preserved in the Metropolitan
Cyprus,
from

four

New

Museum

of

York.^

The combinations

of

human

with animal forms, which played

so great a part in the sculpture of

Egypt and Mesopotamia,

were rarely adopted, and never favorites in that of Greece.


in this type of centaur,
It is hence the more remarkable that,
as in the primitive gorgones {fiop/u,o\vKe2a), the archaic art of

Hellas and Etruria even exceeded the malformations in vogue


among the barbarians. Oriental art does not appear to have
ever figured a quadruped with

human

Such a form was,

legs.

however, too monstrous to be long retained by the rapidly

advancing sculpture and painting of the Greeks, and was soon


Entirely apart from its horrid
proscribed as disgraceful.
female
be a forgery of the later Renaissance. A gem in crystal, representing a
centaur of this type drinking from a rhyton, in the National Library at Paris
(Chabouillet, Cafa/. des dunees, No. 16S9), is engraved in A. Bougot, Philostrate
Vancien.

Une

Galerie antique, p. 361.


is given by Helbig (WolfNo. 16 and a large cylinder
published by the same writer in the Bulletiifio,

bronze vase with a representation of this kind

gang), Ciste Prenestine, Bullettino,


of ivory ornamented with reliefs,

Roma,

is

1866, p. 144,

1874, p. 210, in an article entitled Scavi di Chiusi.


is now in the Terrosi collection, Cetona.

This highly interesting cylin-

der

First Report, p. 110, notes

The one published

in

and

Gori (A.

2.

Museum

F.),

Etrtiscum,Y\ox&w'C\7i&, 1737-43.

6l, en65; the other by Braun (Emil), Bronzi Etruschi, Annali, 1836, p.
graved in the Monumenti Iiiediti, Roma, 1S36, vol. ii. pi. 29. Compare the referIt is apparently the latter of these
ences in the Btdlettiuo for 1835 =i"d for 1836.
pi.

which reference was made by Helbig, in a paper read at an archaeological


case,
meeting and reported in the BuUettiiw for 1871, No. IV. if this be not the
the list.
then a third bronze, closely resembling the other two, is to be added to
human
3 See Note on a Terra-Cotta FigJirine from Cyprus of a Centaur with

to

Fore Legs, by
America,

I.,

Thomas W. Ludlow,

of the Archaological
photograph of the object.

in Bulletin

Jan., 1883, Boston; with a

Institute

of

ARCH^OLOGJCAL INSTITUTE.

148

was

nature, this combination

lending

itself to

at a disadvantage,

a state of rest, the dissimilarity of the legs

manageable and ludicrous,

human members being


The awkward sprawl of
Assos

the

mode

In

might not be found

any vigorous movement became

intolerable, but

because not

the exigencies of artistic representation.

at

once un-

of locomotion of the

entirely unlike that of

the equine.

the human-legged centaurs upon the

epistyle, for instance,

must have formed a striking con-

Thus the

trast to the easy gallop of their neighbors.

primi-

conformation of the centaur, in which a horse's trunk and

tive

hind legs were attached as an outgrowth to the complete

body

of a

human

being, was given up after but few experi-

ments. The more perfect structure seems to have come into


general use some time before the building of the Assos
as early, at all events, as the time of Pindar, who
temple,
refers to centaurs as horse-legged, " from their dam inheriting

the parts below, from their sire the parts above."

In

fact,

the earlier conception, which in

monumental stone-carving

represented only by our

ultimately

albeit

somewhat

Planudes

of

relief,

Greek mind, as

foreign to the

is

the Anthology

'Avbpodfv (KKe^vd' imros

'

avtbpafie b 'nnrodev dv^p,

KfC{)aXi]S

8 urep aloXos

Imros fpevyerai uvSpa, dvrjp S anoTtephiTai

ittttos

'

ittttov.^

even more directly from another


"Ittttos fi]v aKaprjvos, dvrj

Of ye
^

became altogether

evident from a forcible,

coarse, epigram preserved

dvTjp v6cr(f>i TToBav,

And

is

(f)v(Tis

p S

arfXetrros eKetro

nai^ovaa 60a iffKevrpLaev Imva?


Miiller [Archdologie dcr Kiinst, ed. 1S7S,

Pindar, Pythia, II. SS, ed. Ileyne.

389) seems to place too late a date for this transformation of the human-legged
to the horse-legged centaur, stating that it took place " etwa seit Pheidias."
Schmitz (Article Centauri, in Smith's Dictionary 0/ G nek and Roman Bio\;raphy

and

MytJwloi^y,

London, 1876) likewise

states that the latter

form "was probably

not used before the time of Phidias and Alcamenes."


2

Anth. Palat., Append. Planud., 115.

^ Ibid.,

116.

INVESTIGATIONS
The appearance
Assos

of three

relief suffices

mann/ who

to

AT

ASSOS,

18S3.

49

human-legged centaurs upon the

disprove

assumption

the

of

Kliig-

has argued that this form was not that in which

the archaic designers figured to themselves the entire race


of centaurs, but

was

a distinction with

which the more hu-

mane and mild-mannered among them were

alone honored

Cheiron, and occasionally Pholos, being thus anthropomor-

must be admitted that the human


members were retained until a somewhat later date in the
phized, as

it

were.

Still it

case of Cheiron, who, from the nature of those mythological

scenes in which he appears, was not

commonly represented

in violent action.^

While the Herakles relief is unquestionably the more charand interesting work, the block newly discovered
The
displays a great advance in respect to technical ability.
acteristic

bodies of the horse-footed centaurs are

formed

the curves of back and belly

tion of the living animal,

much more

show

correctly

a direct observa-

and contrast strongly with the lank

and almost cylindrical bodies of their human-footed neighbors.


The action of the hind legs, though entirely conventional, is

more

true to nature

and the same may be said of the human

trunks, which are better proportioned, and modelled with a

greater understanding of the muscular development.

remarks apply also to the arms,

more archaic
^

relief.

The heads

much

are of a like type,

Kliigmann (Adolph), Sulla Maniera di Rappresentare

tino, 1876, p. 140.

tradition

It is interesting to

These

too short in the

the hair

Centauri, Bullet-

note in this connection that the ancient

Aelian, Var. Hist., IX. 16) represented Mares, the centaur of the Auso-

been human-legged.
Almost all those vases of later date which have been referred to as
showing human-legged centaurs represent the single figure of Cheiron. It is
worthy of especial notice that on the Fran9ois vase (Braun, Emil, Vaso di
nes, to have
2

Clitia

vol.

ed Ergotimo, Atwali, 1848, engraved

iv.

pis.

horse-legged.

54, 55)

Cheiron has human

in the Montitncnti Inediti, 1S44-48,

legs,

while

all

the other centaurs are

ARCH^OLOGICAL INSTITUTE.

I50

looped above the ears and falling in a thick mass upon the

back

of the neck, while lying so closely

show

as clearly to

round outline.

its

upon the

new

In the

skull itself
relief

they

are smaller and in better proportion with the figures, but

much

expressive, being almost entirely free from that

less

uncouth and goblin-like aspect which

is

so attractive in the

centaurs of the corner block, because so well in keeping with


the wild nature of this mountain roaming race, infuriated by

the odor of the wine which Pholos had broached for the hero.

The most

and

interesting,

most important,

of all the

in scientific respects

Assos

reliefs,

by

far the

namely, that represent-

ing Herakles, Pholos, and the three human-legged centaurs


(Fig. 37),

and

ion,

fell

to the share of the explorers in the official divis-

now

is

main features
was stated

regula,

Hence
of a

its

it

of the

was concluded

of the

human

relief,

being.^

to

be hardly possible that the body

was consequently not a centaur, but

In conformity with this view the figure in

First Report, pp. 107-111, PI. 15.

curious argument
i,

original length.

and that the figure standing imme-

July

its

have been sculptured upon the missing

Critic,

intercolumniations, this block

upper surface preserved in

diately behind Herakles

The

to in the Prelim-

judging from the position of the middle

that,

horse could

portion

of Boston, No. S. 11 57.

work have been referred

and the width

upon

is,

Museum

the

In one respect that description was at fault.

inary Report.^
It

in

of this

is

advanced by an anonymous writer

1S82, in respect to this figure.

The author

in the

A^ew York

of a review of the

Report states that there exist ancient representations of centaurs which


as quadrupeds, but as perfect human beings with the sole addiHence, it is argued, the individual standing behind Hertion of a horse's tail.
akles on our relief, although destitute of a horse's trunk and hind legs, may
nevertheless be held to be a centaur, and in fact Pholos himself, who is " thus
First

show them not

represented to distinguish him from the other centaurs." It is scarcely necessary


to enter into a criticism of such a confusion between the clearly differentiated

forms of satyrs and of centaurs.

INVESTIGATIONS

AT

ASSOS,

1SS3.

151

question was identified as lolaos, the companion of the hero,

who

not infrequently appears in ancient representations of

this scene.

calculation of the original length of the blocks,

with reference to the positions occupied by the separate


liefs

upon the

certainty,

has

at first adopted.

members

be determined with

shown the incorrectness

of the identification

Owing

to the excessive irregularity of all the

of the entablature, no estimate of dimensions can

pretend to greater accuracy than that which


decimeters, but even this

in

To

may be

expressed

sufficient to furnish a proof.

is

ascertain the width of the intercolumniation, from centre

to centre of the

we have

columns above which the block was placed

add

to

to

the total length of the two fragments

found the length of the half-regula which


the left-hand

minimum

of

side.

2.7

The

m. makes

it

missing upon

is

thus obtained

result

namely,

which alone are

of so great a dimension.

Now

which the two fragments were discovered

southern corner of the eastern front


to

leaves

which of the four corner intercolumniations

signed.

The

unearthed

is

in the First

evident that this relief was

above one of the corner intercolumniations of the

in

re-

in this case to

epistyle,

exact spot in which

close
little
it

front,

the position

is

to

the

doubt as
to

be as-

the Herakles relief was

indicated upon the plan of the Acropolis given

Report.^

It

may be here remarked

that those

sculptured epistyle blocks which were incorporated into the

rude masonry which surrounded the temple plan seem never


to

have been moved, after their

be assumed
be proved
position

in the
in

This

fact, naturally to

absence of contrary evidence,

the case of

all

upon the entablature

siderations

from the front to the

fall,

rear of the building, or vice versa.

is

actually to

those other reliefs of which the


is

recognizable from other con-

namely, the two fragments of the sphinxes


1

First Report, PI.

2.

B.

now

ARCHAEOLOGICAL INSTITUTE.

152
in Boston,

and the horse-legged centaurs

and the two fragments


boar from the western
relief the

of the eastern front,

of the other sphinxes,


front.

and the Hon and

In the case of the Herakles

conclusion drawn from the place of discovery

is

confirmed by the general direction of the composition, which

would naturally have been advanced towards the centre of


the fagade, rather than towards the sides of the building.

The movement from

left

would thus indicate the

to right, so decidedly
relief to

pronounced,

have occupied either the

northern corner of the rear, or this southern corner of the

To

front.

contain the

full

body

of a centaur

of the individual standing next to

behind the trunk

Herakles would, indeed,

re-

quire the panel to have had a length of about three meters,

greater by 20 cm. than any intercolumniation in the building,

so that the

identification set forth in the First

not without a semblance of reason.


entirely removed,

The

and a further argument

of the relief gained,

by the recognition

difficulty

to

Report was
is,

however,

prove the position

of the fact that the

corner epistyle was lengthened beyond the

axis of the cor-

ner column by one half the thickness of the entablature, so


that there

right-hand
It.

must have been a length


side of this relief to the

of three meters from the

corner of the building.

does not, of course, follow that the

lintel itself

was

of this

length, for the corner epistyle blocks were not mitred, but

overlapped, and, as the lap and true corner seem, as will be

subsequently explained, to have been cut upon the beams


of the side,

we have

to

deduct from the given

total

the

width of these overlaps, which, judging from the lower thickness of the epistyle beams, and the length of the two other

corner beams, was planned to equal somewhat less than one


half of

the corner

regula,

or

about 20 cm.

The

original

length of the relief representing Herakles and the centaurs

may

thus be asserted to have been very nearly 2.8 m.,

or,

INVESTIGATIONS

AT

ASSOS,

1SS3.

153

words, of just that dimension which, taken together

in other

with the overlap, would be necessary to contain the body of


a centaur behind the hero.

Preliminary Report

whose presence

is

is

figure called lolaos in the


all

doubt, that of Pholos,

scarcely less necessary for the identification

Herakles himself.

of the scene than that of

Pholos

The

hence, beyond

hand the drinking

holds in his left

vessel which, in

come down

the most ancient reference to the story that has


us, the friendly

centaur

2KvrTcf)ftov Be
n'lfi'

to

described as handing to his guest

is

Xa/3wv SeVay (^fieTpov

(TTiaxofxevoi, to

^d

01 irapeOijKe

ojy

TpiXuyvvov,

4-6Xos Kepdaas-^

Stesichoros in Athenaios, XI. 499 B. The form of drinking vessel illusJahn (Otto), Beschreibimg der Vasensa7nmlung Kdiiig Ludwig's in der

trated by

(re-engraved and more geni. 6,


Guhl and Koner, Das Leben der Griechen und Rbmer, ed. 3,
Berlin, 1872, fig. 198. 4,) as a skyphos, is precisely like that held by Pholos in
This form of drinking vessel is termed by Pathe relief discovered at Assos.
Pi)iakothek zu Alunchen, Miinchen, 1854, pi.
erally accessible in

nofka (Theodor), Rechcrches


pi. 4,

stir les veritables

a kotyle; but that this identification

the testimony of Athenaios (XI. 478 B),

Stephani,

has but a single handle.

when

describing the skyphos as

is

Noms

des

incorrect

who

Vases Grecs, Paris, 1829,

may be

plainly seen from

expressly states that the kotyle

Compte Rendu, 1873,

a henkelloses Gefdss,

for

^^

likewise in fault

that the skyphos

was provided with handles

is plain from a reference of Simonides (in Athenaios,


XI. 498 E) to an ovaroevTa a-Kixpov. The correct identification of the ancient
name is due to Gerhard (Eduard), Intorno le Forme di Vast Vokenti, Annali,
1831, p. 257, and Montimenti Inediti, 1831, pi. xxvii. 46-49; also, Ultime

Forme di Vasi Greci, Annali, 1836, pi. c. 24, 25, and 47, and BerAntike Bildwerke, Berlin, 1836, Beilage A, No. 28. Compare the critical
remarks upon this point by Letronne (Jean Antoine), Observations Philologiques
Ricerche sidle
lin's

et

Archeologiqnes sur

les

Noms

des Vases grecs, Paris, 1833.

The skyphos, a homely substitute for the kantharos, seems to have been particularly in use among country people.
Thus Asklepiades of Myrlea (in Athenaios, XI. 49S F) says, " None of those who live in towns, not even citizens who
are but moderately well off, use the skyphos, which is employed only by
swineherds and shepherds, and men in the fields generally." Alkman (in Athenaios, XI. 499 A) speaks of a huge skyphos " such as is owned by shepherds "

and Eumaios
Theokritos

off"ers

wine

[Idyll., I. 143)

to Odysseus in a cup of this kind {Odyss., XIV. 114).


even uses the word for wooden milk pails: evidently

such two-handled vessels as are

still

employed by the herdsmen of Sicily and


in which the milk is

Calabria in dipping out whey from the enormous caldrons

ARCH^OLOGICAL INSTITUTE.

154

The

left

hand

is

raised as

if

deprecation of so rude an

in

The

interruption of his hospitality.

quiet and almost stately

attitude of Pholos contrasts strongly with the wild gestures of

the other centaurs.

His head

is

of a higher type than theirs,

He

being smaller and better formed.


others, but his beard

which

shown

is

shorter and

is

bearded like the

The

more comely.

and well formed

chest,

and
though the head, supported upon too short a neck, droops
is

in direct profile, is full

slightly forward, as

to

if

indicate the physical weakness of

aged centaur, the carriage of the shoulders

this

In

dignified.

fine,

nary Report.

figures

is

erect and

the endeavor of the sculptor to give a

certain nobility to this personage

The other

clearly apparent.

is

were correctly identified

Hence

in the Prelimi-

the following remarks concerning

them

should be taken in connection with the general description of


the

relief,

and of the scene which

it

represents, given in that

volume.^

The

chief attention of the sculptor

was evidently devoted

to

the figure of Herakles, which displays a closer observation of


nature, and greater care in execution, than do the centaurs.

In spite of the surface weathering, the details of the head are


distinct,

still

and prove how firmly the outlines must

nally have been marked.

Though

elevated considerably

origi-

more

supposed to be in
and is fitly representative of the
rude hospitality which Pholos offered to the Doric hero.
To this it may be added that the skyphos had come to be peculiarly identified
with the gluttonous Herakles, who was said to have originally used this kind of
a cup while on his expeditions (Athenaios, XI. 500 A). Macrobius [Sat., V. 21)
"To drink the cup of Herakles" evisays, " Scyphos Herculis poculum est."
boiled.

In short, the vase

the hands of the

dently

came

to

more

mean

is

precisely such a one as might be

civilized of the centaurs,

excessively large potations.

(Plutarch,

Ahx.

75.)

Com-

pare Virgil [Aen., VIII. 27S) and the commentary of Servius on this passage.
Lucian (Conviv., 14) particularly refers to the position in which the ancient

Greek painters were wont to represent Herakles, drinking


and holding this cup in his right hand.
1

First Report, pp. 108-110.

in the

cave of Pholos,

INVESTIGATIONS AT ASSOS,

1883.

55

than seven meters above the eye of the beholder, every feature
must have been readily distinguishable, even without the

emphasis of

color.

The

hair,

cropped as that of Herakles


is

though by no means so closely

is

usually

shown

in later times,

comparatively short, as becomes an athletic hero.

of clustering curls, the separate locks of

Masses

which were without

doubt represented by painted spirals, are indicated by a broad


Between these
welt above the brow, and by a short chignon.
a sharply pronounced curve displays the outline of the skull,
in fact of the same hypsibrachycephalic
high and short,

type as the crania of the ancient Assians themselves. The


The receding forehead is
features are those of a young man.

higher and somewhat more convex than that customary in


the more advanced style, forming with the coarse and promi-

nent nose a profile resembling that of the heads of such statues of athletes as the so-called Apollos of Thera and Tenea-

The

eye, standing in a slightly oblique position,

almond-shaped, almost as

if

drawn de

face.

is

The

full

and

lips

are

thick and pouting, nearly touching the lower surface of the


nostrils

the corners of the

mouth

are

still

drawn upwards,
inferior jaw

The

but the archaic smile has almost vanished.


In short, the face
the chin round.
is massive

is

that of a

vigorous, unintellectual athlete, excellently characterized.

Although the energetic movement of the body has evidently


been studied from the living model, and is rendered with considerable freedom and technical skill, it is in the forms of the
trunk and lower limbs, rather than

in

the head, that

we meet

is unwith distinct reminiscences of archaism. Thus


the buttocks are too small, and yet too
naturally compressed
part
of the legs is of too convex a curve
protruding the upper

the waist

upon the front side; the knees, especially the left, are insuffiThe feet
ciently indicated, and too much rounded in outline.
are small, the heel and ankle having but

little

projection,

ARCHAEOLOGICAL INSTITUTE.

156

while the toes are too long and too

Now

flat.

all

these fea-

tures are characteristic of black-figured vases of a style which,

without taking into account the influence of provincial backwardness, would be ascribed to the end of the sixth, or

On

decade of the fifth century before Christ.


a highly intelligent observation of nature
in those parts

which are

which no

model

of

set

in action,

of forms

employed by the designer of

and

is

in

first

the other hand,

no

less noticeable

the representation

and proportions can have been


this figure.

Indeed, the treat-

ment is here extremely skilful. The swift yet cautious stride


upon the slippery ground ^ the inclination of the trunk,
;

thrown forward
as to hold the

in pursuit

bow

the outstretching of the arms, so

entirely free

the slight lowering of the

head, in order to take sight of the arrow;

ments are

clearly expressed, while the

perfect equilibrium.

The muscles

body

all

is

of the right

distended with the effort of drawing the stout

these move-

brought into

upper arm are

bow

the

arm is stiffened in full resistance. The chest, expanded


by a deep breath, is excessively thick, this effect being
duced by a greater exaggeration

of the

left

as

if

pro-

muscles of the back

This peculiar formation evidently

than of those of the breast.

resulted from an attempt to indicate the displacement of the


right shoulder

by the

into account, the

strain of the arm, but

even taking this

back appears too round, and even slightly

The difference in plane between the lower ribs and


abdomen is so marked as to cause the latter to appear

humped.
the

unnaturally contracted.

trunk
hero,
1

is,

This extreme development of the

however, entirely in accord with the character of the

who was conceived by

the Greek sculptors of

all

ages

Nephele, the cloud mother of the centaurs, had during the combat deluged

the earth with torrents of rain, so that Herakles could hardly stand upright upon
the slippery ground, while his four-footed opponents were not thereby discomforted.
tail of

Diodoros (IV.
the legend.

12. 6)

makes

particular mention of this picturesque de-

AT ASSOS,

INVESTIGATIOiVS

1SS3.

57

strength and energy,


endowed with supernatural powers.
It has been truly remarked that the testimony of Greek
writers^ clearly shows us that what seems an undue exag-

human being

as

of

extraordinary

rather than as a demigod

geration in such figures as these

was but the emphasized

form as

idealization of the athletic

it

appeared

in reality.

bow

Herakles attacks the affrighted centaurs with the

His peculiar

closely connected with several of his exploits.

weapon was renowned

the use of this

skill in

so

in

the Homeric

poems,^ and was, throughout antiquity, so striking a charac-

'O/jLijpiKT} (TTokrj

As

of the

accoutrement

this

from

his further

first

Preller

as his

has ingeniously surmised, the

Herakles of the poet Xanthos

signifies

an archer, as distinguished

of the hero as

equipment with a

club, which,

if

we

are to be-

statement of Megakleides, preserved by Athenaios,^

lieve the

was

bow must be regarded

hero that the

of the

teristic

original attribute.

introduced by Stesichoros. Thus the reference of Pau-

sanias^ to the primitive crxw'^ o^ Herakles upon the chest of

Kypselos in Olympia asserts the hero to have there appeared

bowman,

as a

of the

same

type, doubtless, as the 'H/ja/cX?}?

For instance, Aristophanes, Clouds, 1009-1014.


was with the bow that Herakles wounded Aides, and even Hera herself
Compare also the renown of Herakles as an excellent bowman in
(//., V. 395).
the Odyssey (VHI. 224 and XI. 606). It is quite possible that the last mentioned
1

2 It

of these passages, relating to the descent of Herakles to the infernal regions,

is

the interpolation of a later age (compare the special literature of this question, in
particular Lauer, Quaestiones Home^icae, Berolini, 1843)

recognize

in

who

tions obtaining in regard to the hero,

bow, and an arrow


shaft."

ages;

The
e. g.

but we

may

nevertheless

the tradition there recorded a true exponent of the primitive concep-

at the string,

is

described as " armed with a naked

always like unto one about to

peculiar skill of Herakles in archery

Euripides, Here. Fur., 157-164, 188,

was celebrated also

Preller (Ludwig), Griechische Mythologie, Leipzig, 1875, ' ^29, n.

*
^

Xanthos in Athenaios, XII. 512 F.


Ibid.
For the costume of Herakles see

Pausanias, V.

Miiller's Dorians, B.

ii.

in later

etc.

17. 11.

let fly

ch. 12,

i.

ARCH^OLOGICAL INSTITUTE.

158

combat with a human-legged centaur, shown upon


the archaic bronze rehef before mentioned as having been
To^oTt]^ in

recently unearthed at Olympia, or as the Herakles of the

same subject, now under consideration.


By its contrary flexurC; we recognize the weapon which he here
holds in his hand to be that given to him by the Scythian

Assos

relief of

the

shepherd Teutaros,

earliest instructor of the

archery,^ a legend

significant,

Oriental origin of

that

in regard to the

all

bow and

it

may

was known
its

Dorian hero

be surmised, of the

to the primitive

No doubt

fittings.^

in

concerning the peculiar form of this Scythian bow.


curve, re-entering to the bar grasped by the hand,

Greeks

can exist
Its

is

double

compared

shape of the cursive %, as written towards


fifth
century before Christ. Moreover, the
of
the
the close

by Agathon^

to the

ancient geographers
of the

were accustomed

to

compare the outline

Black Sea to that of a Scythian bow

coast, with its

the northern

two great gulfs separated by the promontory

of the Crimea, standing for the

bow

straight southern coast for the string.

itself,

the comparatively

This graceful contrary

1 Scholiast to Theokritos [Idyll, XIII. 56), quoting from Herodoros.


Compare Lykophron, Alex., 56 and 45S, with the commentary of Tzotzes, and also on

V. 50.
- The Oriental origin of Greek archery has been referred to in the previous
chapter, p. 45, in connection with the bronze arrow-heads of Persian shape found
upon the Acropolis of Assos. Kaoul-Rochette CDishe), Sur FHercule assyrien

dans ses Rapports avec V Horitle grec, [Mcmoires de V Academic des Inscriptions, Paris, 1848, vol. xvii., deuxieme partie,) has called attenet phcnickn, considh'e

between the 'HpacA.7Js toI^ttjs of the archaic coins


and the types of the royal archer upon the darics of Persia.
3 Agathon, in Athenaios, X.
This was, of course, not the later
454 D.
sigma of segmental shape. Euripides and Theodektes (in the same passage of
Athenaios), the one writing in the same age as Agathon, the other nearly a century later, both compare the form of the letter sigma to that of a wavy lock o
The contrary flexure of the letter, and of the bow which it is said to
hair.
tion to the striking similarity
of Thasos,

resemble,
^

is

thus fully assured.

Dionysios Periegetes, 156.

XXII.

8. 10.

Strabo, II.

5.

22, p. 125.

Ammianus

Marcellinus,

INVESTIGATIONS

AT

ASSOS,

1S83.

nature of wood, must have resulted

flexure, so foreign to the

from the employment of the horns of animals

the bases

facture of bows,

159

of

in

the manu-

two horns being attached by

a splicing of a metal ferule to a short, straight middle piece.

Such were the weapons so accurately described by Horner;^

bow thus came, even

while the

in

later ages, to

be termed

Kepa<;?

The Scythian bow was

the favorite

arm

of Herakles,

and

continued to be a characteristic attribute of the national hero


until the latest ages of

remembered,

literature

and

As

art.^

will

and armor-bearer Philoktetes,* which,

friend

his

to

Greek

be

was the very weapon given by Herakles

this

after

conquering the islands near the Troad and expelling from

them the Carian (Lelegian) population,^


the Trojan war by killing Paris.^ And it
that the Aeolic colonists of Assos, the

finally

terminated

not impossible

is

Greek inhabitants of

may have chosen this episode of the Centauromachia from among the many deeds of
Herakles on account of the connection of this invincible bow
the ancient capital of the Leleges,

with the traditional history of the land which .they occupied.

As shown upon

scarcely

our

relief, it is

more than

described in the

Iliad.

IV. 105; Odyssey,

a stout and very short weapon,

half as long as the

So diminutive

XXI.

is

bow
it,

of Pandaros

indeed, that

we

Iliad,

As, for instance, in the passage of Strabo before quoted, and Theokritos,

Idyll.,
8

XXV.

395.

2c6.

reference of the twelfth

attribute of Herakles

Christian century to the Scythian

bow

as an

contained in the commentary of Tzetzes to Lykophron,

is

Alex., 917.
*

Diodoros, IV.

Imag.,

17.

36, 102,

Philostratos, Her., V.

I.

Philostratos the Lemnian,


6.

Hyginus, Fab.,

and other ancient authorities.

* Philostratos,
8

38. 4.

Scholiast to the Iliad, II. 724, ed. Bekker, 90 B,

The

Her., V.

3.

ancient authorities in support of this version of the legend, ranging as

down to Kedrenos, are too numerous, and


known, to be quoted here

they do from Sophokles


least too well

in part at

ARCH^OLOGICAL INSTITUTE.

l6o

was but

that archery

are led to suspect

There

ascribed.

ample space upon the

is

be

to

is

the in-

relief for

troduction of a larger bow, and the careful

by

practised

little

the Assians at the period to which this sculpture

observation of

nature, so evident in the muscular development of the bow-

man, could not otherwise have failed to be extended to this


important adjunct, which thus appears rather as a model,
a plaything, in

short

conventional

ture of the

bow

with

it

is

stretched

Moreover, the arrow

too acute an angle.

represented, as

na-

horns do not taper sufficiently, and the string

much

as

the

shown in the thickening of the tips by


wound around them but the curved

is

the ends of the cord

to

than

attribute,

Some acquaintance

really effective weapon.

should have been, with a

bow

not

is

thus bent in

actual use.

The

fact that

Herakles

lion's skin is exceptional

is

depicted at Assos without the

among

archaic works of

art,

there

being but very few instances of this guise upon black-figured


vases,^

and may perhaps be adduced

that the date of the temple

is

the close of the Persian war.

the

same

effect is to

and beardless type

of

Herakles a very
to

support of the belief

more

forcible

argument

be based upon the adoption

ful

Assos are

in

be placed at least as late as

to

in this figure.

d*efinite indication

be assigned to the

before Christ, rather than to a

We

have

of a

in the

more remote date;

upon black-figured vases belonging with certainty


is

for,

One

is

almost invariably shown

The hero

which

shown by Welcker (Friedrich Gottlieb), Rappresentazioiii delP Idra


also
p. 103, and Monumenti Inediti, pi. 46, Roma, 1836

Lernea, Annali, 1842,


in the Alte

while

to the sixth

as bearded, he has here the beardless youthful form in

head

that the sculptures of


half of the fifih century

first

century before Christ, Herakles

to

youth-

Denkmdler

of the

attacking the hydra

is

same author,

vol.

iii.

pi.

6,

Gottingen, 1849-64.

armed only with quiver and sword.

INVESTIGATIONS AT ASSOS,

1883.

i6l

he appears in the gable group of the temple of Aigina and in


the metopes of the Parthenon,
stylo,

works which,

in respect to

must certainly have been in advance of the sculptures of


This manner of depicting Herakles may

provincial Assos.

have been introduced by the Argive sculptor Ageladas, whose


youthful Herakles, preserved in Aigion,
sanias

manner which seems

in a

to

is

show

mentioned by Pauthat the represen-

tation of the hero without a beard

was an innovation

age when that

The determination

artist

was

at work.

in

the

of the

Brunn has remarked in his discussion of this point,^ one of the most difficult questions in
the history of ancient art.
For our present purpose, it will,
however, suffice to bear in mind that Ageladas, having been

exact date of Ageladas

is,

as

alive at least as late as the eighty-second

Olympiad, cannot

assumed to have created this type of Herakles, which


subsequently became common, before the termination of the
well be

Persian wars.
In regard to the comparative iconography of the Assos
relief,

a striking parallel presents itself in the well-known

Karapanos

of

relief

Herakles drawing the bow.

Attention

has been called to the similarity of these figures by Emer-

who has engraved them side by side for the purpose of


comparison. The genuineness of the Karapanos relief, asson,3

signed by Rayet

to the first years of the fifth

Christ, has been questioned

conceded

work

century before

by Emerson, who nevertheless

have retained many features of some


original very similar to the Assos sculpture.
Furtwangler,^
this

to

Pausanias, VII.

Brunn, Griechische Kiinstler, 1857,

Emerson

vol.
*
6

i.

24. 4.

(Alfred),

p. 152, pi. 5.

Two Modern

i.

64.

Antiques.

American Journal ofArcJi(Eology

Baltimore, 1885.

Rayet (Olivier), Momiments de PArt Antique, pi. 23. Paris, 18S0-84.


Furtwangler (Adolf), American Journal of Archceology, vol. ii.

Baltimore, 1886.
II

p.

52.

ARCH^OLOGICAL INSTITUTE.

62

on the other hand, believes the Karapanos relief to be, not a


modern, but an ancient imitation, ranking it among the finest
known examples of archaistic art, and suggesting the first
century before Christ as the probable date of its execution.
Furtwangler bases his argument wholly upon the style of
the

He

relief.

size

compares

it

with another slab of the same

(showing Herakles with his knee upon the neck of a

which he holds

stag),

have belonged

to

representations of the labors of


this view,

which asserts the Karapanos

may be

but archaistic work,

the similarity between


style

is

it

and the sculpture

panos

relief

at

be an antique

fact, that,

of the

the assumption

the time of the

is

series of

direct proof of

relief to

resemblance, no other figure of this type


existence, and

derived from the

so great as to exclude

same

to the

Herakles.

while

Assos
of

known

epi-

chance

to

be in

discovery of the Kara-

the Assian sculpture was

still

buried beneath the

earth.
It

is

scarcely possible to assume the archaistic relief to

been imitated directly from the decorations of the


temple of provincial Assos, especially as it appears to be but

have

one of a

We

series of representations of the labors of Herakles.

have hence to seek

for

some common

original from

the leading features of both these works were derived.

which
Furt-

wangler suggests that these archaic originals were statuary


groups of the series of Kritios and Nesiotes, inasmuch as the
head of the Karapanos Herakles bears a close resemblance
to that of

Naples,

Harmodios
itself a

dedicated in 477

in

the group of the Tyrannicides at

copy of the work by Kritios and Nesiotes,


b. c.

by the Athenians.^

This striking

re-

semblance, to which attention was called by Rayet, has been


1 Marmor Parium,
C. I. G., vol. ii. No. 2374, Epoch,
Eergk (Theodor), Zttr Periegese der Akropolis von Athen.

thumswissenschaft, vol.

iii.

p. 972.

Giessen, 1845.

i.

line 70.

Zeitschrift

Compare
fur Alter-

INVESTIGATIOXS

AT

ASSOS,

1SS3.

163

When

recognized by both Furtwanglcr and Emerson.

fully

the

critics of opinions so various

panos

age when Greece was a

to the

ascribing

it

and the

third

unanimous

it is

Roman

certainly

it

the Kara-

province,

to avoid the conclusion

difificult

that the original from which


is

in

modern forgery
are so
upon the type from which it must have

even terming

in fixing

been imitated,

rived

one seeing

a work antedating the Persian wars, the second

relief

these characteristics were de-

indeed to be referred

the school

to

in

question.

Moreover, the peculiarities of the style of the Kritios and


Nesiotes, as described by Lucian,^ (a most competent observer,

himself trained as a sculptor,) are certainly recognizable in

both the Assos and Karapanos


Plerakles

of

tightly

drawn

vevpoiSr]

lithe

in,

compressed

We

reliefs.

a clear illustration

of

have

in the figure

the term uTreacpijfjLeva,

like the

abdomen

of our hero

Koi aK\r}pd, sinewy and rigid, of firm rather than

and supple muscles

and

the execution of the entire

in

that sharply cut composition of the outlines

relief,

what exaggerated emphasis

of the physical

and some-

development

result-

ing from a too distinct demarcation of the protruding muscles,

which

We

is

referred to as d/cpc/Bw^; diroTeraixeva

rah

ypa/xfjcais.

thus have good grounds for the belief that the style of

now under consideration was influenced by


and Nesiotes, and may perhaps even go so far

the Assos relief


that of Kritios
as to

assume that the type of the Herakles which here ap-

pears was a direct creation of these sculptors.

The converse

cannot be admitted for a moment.

It is

to entertain the supposition that the

work of Athenian

obviously impossible
artists,

who

represented the highest contemporary development in


the modelling of the human figure, can have been in any

way

influenced by the rude decorations of a building in a


provincial town of Asia Minor.
1

Lucian,

/i^Aei.

Praec.

9.

ARCH^OLOGICAL INSTITUTE.

164

This identification

is

entirely in

harmony with

which other considerations would lead us


temple of Assos.

the

If

building

was

the date

to assign to the

reality

in

erected

during the age immediately succeeding the Persian wars,


let

us say between the seventy-fifth and eighty-fifth

piads,

nothing would have

Olymbeen more natural than that its

sculptures should have displayed in the better figures

some

traces of the contemporary art of Athens, with which city

Assos was then

politically allied,

and to which she must have

looked as the great leader of intellectual and

Even

artistic

advance.

as the plan of the temple was copied almost exactly

from that
sculptured

of

the

Theseion, the types of

were

decorations

works of the Attic

derived

relief

thus supplies us with an indi-

cation as to the artistic style of the


figure of Herakles

contemporary

school.

While the Karapanos

that this original

most perfect

its

from

was derived, other

was a

relief of

work from which


parallels

make

it

this

plain

considerable extent, depicting

the combat of Herakles and the centaurs of

Mount Pholoe

in

the same general composition as that which appears in the

sculptured epistyle found at Assos.

There

exist a

number

of

in

which the

grouping of the figures evidently was determined,

in greater

painted vases with representations of this scene

or less measure, by reminiscences of


art

known throughout

some archaic work

of

the Hellenic world, and regarded as

The composition and forms

typical of the subject.

of this

popular original were imitated, not only in the monumental


decorations of a provincial temple, but in the paintings of
those exquisite vases which were to be found in the dwellings
of every

such

Greek

as, to

citizen

take an

even as to-day a popular picture

example, Rubens's Trinity

is

not only

copied in the altar-pieces of the churches of small towns, but


is

more

or less recognizable in the coarse prints

which

in

ARCH^OLOGICAL INSTITUTE.

66

countries

Catholic

enliven the walls

of

most humble

the

cottages.

Chief

among

the vases of this class

is

an archaic amphora,

probably of Attic manufacture, found at Caere, and


the

Museum

Assos
great

are

relief

Upon

of Berlin. ^

now

in

the chief figures of the

it

So

readily to be distinguished (Fig. 38).

the similarity that no doubt can obtain in regard

is

to the fact that

both of these representations were influenced

The

by some common model.

composition, advancing from

almost exactly the same.

left to right, is

Herakles, bending

the Scythian bow, strides with body bent forward, the ad-

upon the ground, while the right is parThe


tially raised, as in the Assos and Karapanos reliefs.
retreating centaurs are in the same peculiar position as those
their front legs
shown upon the newly discovered block,

vanced

left foot flat

overlapping, while their hind legs are behind those of their

In general outline the likeness

neighbors.

to establish the point in question


this kind,

it

is

in

is

quite sufficient

a comparison of

obviously the similarities, and not the differ-

ences, of design

which require

as proof of a relation to

features

for,

to

be taken into consideration

some common model.

When

these

of similarity surpass the narrow limits of chance

resemblance, the fact of some imitation, conscious or unconscious,

is

at

once fully established.

This remains true,

may be the variations in the treatment of


dependent, it may be, upon the exigencies of the

whatever

adapted, or upon the

detail,

space to
individual

which the composition

is

taste of the designer.

In consideration of the further fact

that these representations are not supposed to have been deGerhard (Eduard), Auserlesene Griechische Vasenbilder, Berlin, 1839-5S,
The tracing reproduced above (Fig. 38) was made by me from
pi. 119.
the vase, for the purpose of this illustration, as the lithograph given by Gerhard
1

vol.

is

ii.

incorrect in certain details.

INVESTIGATIONS

AT ASSOS,

1883.

rived the one from the other, but from a third

167

and

still

more

archaic composition, the general agreement must certainly be

admitted to be surprisingly close.

Another painted vase, representing this scene, and appasome of its features from the common prototype, is that found at Akrai, and published by Judica.^
The
figures of Herakles and his antagonists are here shown in
rently deriving

much

the

same arrangement, the composition being likewise

from

left

to

arm

One

right.

here preserved,

is

detail,

Pholos

uplifted, as at Assos.

the Caere vase,

altered in

standing behind the hero, with

Other vase paintings of

would doubtless be found on examination of

all

this

type

those repre-

senting the subject, few of which have been published in


plates, or are distinctly recognizable

from the short descrip-

tions given in the catalogues of the various collections.^


1

Judica (Gabriele), Le Antichita di Acre, Messina, 1S19.

As was mentioned

illustrations of the

in

the First Report (p. loS, note

i),

seventeen antique

combat of Heralcles with the centaurs of Mount Pholoe have

been quoted by Stephani, Cimipte Rendu, 1873.

This

list

has been unquestioningly

referred to as correct by both Colvin and Puchstein, in the


p. 142,

note

2,

and

however, make

it

p. 145,

note

evident that

r.

many

An
of

works quoted above,


examination of these representations will,

them

Thus, the most important

are, in reality, not at all referable to

monumental respects, the sculpture


upon the Roman sarcophagus published by Braun (Emil), Sarcophago rappresentante Combattime7tto tra Ercole e Centauri, Monumenti Inediti, 1855, pi. 19, which
figures as No. 15 of Stephani's list, shows one of the centaurs to have seized upon
a woman, one of whose feet and some folds of whose drapery, visible upon the
shattered side of the coffer, were evidently overlooked by the learned Russian
the subject.

archaeologist.

The rape

in

thus indicated cannot possibly be brought into connection

with the combat of Herakles with the centaurs of

Mount Pholoe, who were on

occasion attracted solely by the odor of the liquor.

It is

tor of the sarcophagus has illustrated an episode from the

one of the most popular subjects of ancient

art.

that

probable that the sculp-

wedding of Peirithoos,

The same

objection

is

appli-

cable to the bronze of Antoninus Pius,^hich Stephani cites as his sixteenth


stance.
pi.

(Published by Foy-Vaillant (Jean), Selectiora JVumismata,

25; Beger (Laurentius), Hercules Etlmicorum, Berlin, 1705,

Grandmaison (Aubin Louis), Galerie Mythologique,

pi.

in-

Parisiis, 1694,

18; Millin de

No. 437 ;
Guigniault (Joseph Daniel), Religio7is de VAtttiquite, Paris, 1825-51, pi. 170, No.
659 ; Cohen (Henri), Description Historique des Monnaies /rappees sous r Empire
Paris, 181

1, pi.

195,

ARCH^OLOGICAL INSTITUTE.

68

By

putting these various indications together

we can

gain

a tolerably definite idea of the characteristics of that archaic

work

now
The

of art,

imitations.

whose fame

lost,

general

outlines

is

many

attested by so

and

composition,

of its

some of the peculiarities of its style, are clearly recognizable,


and have been set forth at sufficient length. In other respects we can only assert that the original representation

much

longer than broad, and was, in all


composed within the limits of some architectural framework.
Ancient writers have left us no account
of any works of Kritios and Nesiotes representing Herakles
and Furtwangler's identification of the Karapanos relief, as

occupied a

field

probability, a relief

derived from

" the

series " of those artists, should,

without

doubt, read a series.

celebrated work in

representing the labors of

relief,

Herakles, in an architectural framework, and fully answering


Romain, Paris, 1859-68,
carrying off a

woman,

vol.

ii.

p. 338,

perhaps

No.

436.)

Homados

Here

also a centaur

is

shown

with Alkyone, the sister of Eurys-

by Guigniault. In this case the objection has been anticimeans fully met, by Stephani, who errs also in describing the
With the representations of the wedding of Peirithoos we have,

theus, as identified

pated, but by no

coin as silver.

furthermore, to class the red-figured vase published by Inghirami (Francesco),


Pittiu-e di

Vasi Eiruschi, Firenze, 1852-56, 2d

ed., vol.

(Pierre Frangois), Antiquites Etrusques, Paris, 1785,

No.

The

12.

fine silver vessel

now

in

i.

pi.

pi.

79,

and Hugues

124; Stephani's

list.

Munich, published by Arneth (Joseph

Gold uud Silbcrvionuviente des k. k. Aliinz, und


Wien, 1850, pi. S. II, Stephani, No. 17, also does
not relate in any way to the legend of Herakles and Pholos, merely showing
the struggle of two Lapithae with two centaurs before an image of Ares; and the
vase published by Moses (Henry), A Collection of Antique Vases, Altars, Paterce,
Tripods, Candelabra, Sarcophagi, etc., London, 18 14, pi. i, represents Herakles
between two centaurs, with nothing to identify the site of the combat. Still another of Stephani's references (No. 13, cjpoting Maximis, Mus. Etrusc, vol. ii.
pi. 77) is erroneous, there being no such representation shown upon the plate in
Calasanza von). Die Antiken

Aiitiken Cabineltes in IVien,

question.

Several of the rest are catalogue entries, so vague that the real char-

acter of the scenes depicted


fi-om.
list is

Even taking
to

upon the vases cannot be

fully ascertained there-

these latter into account, as correctly identified, Stephani's

be reduced from seventeen to eleven examples.

AT ASSOS,

INVESTIGATIONS

1SS3.

69

the requirements of the case in point of date, was the sculpture of the temple of

We

Gitiadas.

of

Athena Chalkioikos

not

manded

to

only

those

labors

upon the

which

of his

own

free will.

category, the combat with the centaurs of

scarcely have been lacking.


this sanctuary

struck,

repre-

wall,

Herakles was com-

perform by Eurystheus, but also those exploits

which the hero engaged

in

work

learn from Pausanias^ that these reliefs,

evidently arranged in compartments


sented,

at Sparta, the

The fame

was widely extended

In the latter

Mount Pholoe can

of the decorations of

for instance, coins

were

not in Sparta alone, but in other towns, with the

But
beyond the suggestion that a connection may have existed
between these reliefs and those of Assos, our present entire
type of the sacred effigy preserved within the building.^

ignorance concerning the artistic style of Gitiadas


relation to the

and

its

contemporary work of Attica, does not permit

us to go.
Little

remains to be said concerning the other figures

shown upon
taurs

this relief,

namely, the three human-legged cen-

who hasten away from

the arrows of Herakles.

general description of them has been given in the First Report,

and the chief

been discussed.

peculiarities of their form have already


Although evidently the work of another hand

than the horse-legged centaurs, they display indications of

having been imitated from one and the same model with
them. The arrangement of hair and beard, and of hind legs
and tail, is entirely similar; and, in particular, the position of
the outstretched arms of the middle centaur upon the corner
block closely resembles that of the others, the juncture with

the body showing the


1

Pausanias, III.

same malformation

of

the muscles,

17. 3.

Compare Koner (Wilhelm), Darstellung des Standbildes der Athene Chalkioekos zu Lacedaemon, in Koehne's Zeitschrift fur Miinz, Siegel und Wappe)ikiaide,
2

vol. v., 1845.

ARCHAEOLOGICAL INSTITUTE.

170

hand are turned

while the thumbs of the

left

the right down, in the

same way.

These

up,

more remarkable, because

ao-reement are the

and those of

peculiarities

of

of

the srreat

technical superiority and general correctness of form noticeable in the relief

now

to Herakles, turning

first

back

The centaur

published.

his

head

stone in his uplifted right hand

shoulder a thick club, these

to

aim

nearest

his missile, holds a

the foremost bears upon his

being the weapons with which,

according to the legend, the centaurs of

Mount Pholoe

carried

on the combat.^

Among

the discoveries of the second year

is

the fragment

44 cm. long and 27 cm. high, represent(Fig. 39.)


of
legs
a running centaur.
Being
hind
the
ing
the lower left-hand corner of the slab, it shows a rebate, in
of a metope, about

plan, about 25

which was
joining

the

feet,

to

by

15

mm., cut upon that side of the stone

be hidden behind the projecting edge of the ad-

triglyph.

fillet

and corresponds

of the epistyle blocks.

cm. broad forms a plinth for

to the tainia

upon the lower edge

This emphasis of the architectural

a decided drawback in aesthetic


framework of the metopes
respects, as it cramps the field available for sculptured representations

is

an archaic feature, omitted entirely from the

1 That the centaurs were armed with the branches of trees is attested by
Hesiod {Sait. Here, iSS), Pindar [Frag., 144, ed. Bergk), Apollonios Rhodios
{Argon., I. 64), Orpheus [Argon., 173), Diodoros (IV. 12. 5), Apollodoros (II. 5.
We learn from the three authorities last
4. 3), and Ovid [Meiam., XII. 507).

named

that they also threw stones.

the shield of Eurypylos, described by Quintus of Smyrna [Posthom.,


VI. 273), the centaurs of Mount Pholoe were represented as fighting with clubs,
attacking Herakles. The poet, without doubt following an ancient epic proto-

Upon

type, says, "

others

still

Some were shown

prostrate

upon the pines which they grasped, while

carried on the fight with like weapons."

The names

of two centaurs, written out upon a vase published by Gerhard


und Katnpanische Vasenbilder des Museufns zu Berlin,

(Eduard), Etruskische

Berlin, 1843, pi. 13, viz.

wooden club and

TAAIGS and nETPAIOS,

are evidently derived from the

the rock with which they threaten Herakles.

INVESTIGATIONS
Theseion and Parthenon.

AT

It had,

ASSOS,

1SS3.

171

however, certain practical

advantages, protecting in some measure the raised portions


of the relief

from being spHt away during the process of

ing and setting the block, and

stone of Assos than

it

is

more

lift-

tolerable in the coarse

would have been

in

the marble of

Attica.

The

small portion of the figure which remains

is

perfectly

sharp and free from weathering, and shows the sculpture to


with

have been

executed

much

The

hoofs and

the

conven-

legs

care.

are

in

but some-

tional position,

what

farther

apart

than

any of those represented

upon the other

reliefs

centaurs.

be

It will

of

recol-

lected that one of the three

Fragment of a Metope.
Hind Legs of a Centaur.

Fig. 39.

metopes removed from Assos to the

Louvre

also rep-

resents the single figure of a centaur, galloping, with a club

upon

his shoulder.

The hind

legs

shown upon the newly

dis-

covered fragment are of the same size as those upon the


block in Paris.

It is

hence evident that

not contain a second figure,

this

metope

also did

in this respect differing

disadvantageously from the metopes of the Attic

most

monuments

before mentioned.

The

last of the reliefs

found during the second year which

remains for our consideration

is

a fragment of the heraldic

sphinxes once decorating the western front of the temple.

The

greater part of the other sphinx sculptured upon this

epistyle block

1835,

second

was removed from Assos by the French in

and has since been


fragment, found

preserved

in

by us upon the

the

Louvre.

surface

of

A
the

ARCH^OLOGICAL INSTITUTE.

172
earth,

where

had apparently been seen and drawn by

it

The

Texier, was published in the First Report.^

ment, with which

we

and wings

sphinx upon the left-hand side

of the

pletes the sculptured

are

now concerned,

subject,

the

third frag-

contains the body

now

panel

com-

it

lacking but

a small portion of the tainia, split from the second fragment.


(Fig. 40.)

Before going further,

it

should be explained that the fact

of this relief having been situated above the central inter-

columniation of the western front, and not in the corresponding position of the eastern front, can be determined from the
lengths of the half-regulas cut upon the ends of these blocks,

which, together with those adjoining, must have exactly made

up the

total

As

widths of the triglyphs above them.

been already

set

forth, there

can be

little

has

doubt that the

epistyle block sculptured with the four centaurs, found dur-

ing the second year, was placed next to the

relief of

Herakles

and Pholos, and consequently adjoined one of these sphinx


reliefs

upon the left-hand

Now

side.

the half-regula upon

the right-hand end of the relief of the four centaurs

is

ceptionally long, namely, 33 cm., and cannot have been

ex-

com-

plementary to the corresponding moulding upon the newly


discovered
entirely,

sphinx, which

itself

is

27 cm. long.

agrees

however, with the space remaining for a half-regula

upon the shattered end of the other sphinx


during the
Plate 16.

It

first

We

year,

and published

in

block, discovered

the First Report as

are thus as fully justified in assigning these

heraldic reliefs to the front and rear of the building, respectively, as

we

are in the assumption that the four centaurs

formed a continuation

man

They

Herakles.

of the
will

file

retreating before the bow-

hence be distinguished as the

eastern and western sphinxes.


1

Preliminary Report,

p. 115, pi.

19.

INVESTIGATIONS

ASSOS,

73

of this epistyle

important facade, to which our newly discov-

less

ered fragment appertains, with that of

front of the building.

two

1SS3.

compare the sculpture

interesting to

It is

from the

AT

its

pendant from the

glance at the photogravures of the

reliefs (Figs. 40, 41) will suffice to

show that the

figures

were executed from the same design, but by different hands,


differing widely in technical skill.

The

variations in position

manifestly resulted from the sphinxes being framed in panels


of

unequal dimensions, that of the front measuring less than

2.5 m.,

while

that of the

2.6 m. in length.

The

rear

was considerably more than

sculptor of the latter placed his fig-

ures quite as near to the ends of the block as they were in

the former, the entire difference in length falling between


the heads and

Thus the

breasts of the animals.

fore legs

of the western sphinxes were disproportionately lengthened,

while the angle of their elevation was


creased.

It

correspondingly de-

doubt in conformity with

was without

change of angle that the wings were made


lower upon the back.

this

to lie

somewhat

With these exceptions, the

outlines of

the two reliefs are almost identical.

In modelling, however, the sphinxes of the west are decidedly inferior to those of the east.

Although projecting

quite as far from the background, they yet appear

undefined,

betraying

in

technical respects a

influence of the sphyrelaton style.

fiat

and

more marked

The western

relief is, in-

deed, an especially good example of that clumsiness of form


in

the

masses,

the

protuberances

being of basket-shaped

rather than of oval section, and of that angular and straplike

rendering of the details, so indicative of a practice of

beaten-metal work in which the sculptors of Assos were evidently versed.

The curves

of trunk

and haunch are not so

true to nature as in the eastern sphinxes, while the edges of

the relief are

too thick and

cushion-like

to permit

of the

ARCHAEOLOGICAL INSTITUTE.

174

rotundity of the body being effectively rendered.


fine gradations of light

sufficiently

weak and

from the

are

not
too

prominent and too sharply

thin, the tail too little

In short, the western

curved.

hind legs

the

flank,

the

of the

The thigh-bone does

figures are almost entirely lost.

project

Thus

and shade within the outlines

sphinxes are of a dull and

heavy sleekness, while those of the east are sinewy and


vigorous.

Although the main outlines

substantially identical, there are

before said,

are, as

slight differences in the

still

Even

curves and relative positions of the members.

if

we

suppose the design to have been transferred to the surface

from an original cartoon by a tracing, or other

of the stone

mechanical means, there must have remained a certain scope


for the sculptor to display his taste

and technical

skill.

noticeable deviation in the western sphinx from the forms

of the eastern

is

of the wings.

the decorative scalloping of the outer edges

More important and

straightening and

less successful

flattening of the

outlines

are the

the

of

belly,

through which much of the grace and force of the original


has been

lost.

It

evident that the architect or artistic

is

superintendent of the decoration

of

the

temple was well

acquainted with the relative abilities of the sculptors working under his directions,

those

reliefs

which were

building to the

We
ficient

more

and

to

intelligent

may even venture

assigned

execution of

the

be placed upon the front of the

and

skilful

hands.

the supposition, that the less pro-

sculptor of the western sphinxes was, as

with his
ventional

rival,

an

artist of the

methods.

As

old

compared

school, clinging to con-

has already been pointed out, the

more apparent

work, while

style of the sphyrelaton

is

the wings are scalloped

in conformity with a highly archaic

decorative practice.

Another feature

ing to the same conclusion,

is

of

in his

much

interest, lead-

the shape assigned to the termi-

INVESTIGATIONS

AT

ASSOS,

1SS3.

175

nation of the vertical shaft upon which the sphinxes rest their

upUfted paws, and which marks the centre of the symmetrical


In the relief from the rear of the building this

composition.

termination has the form of an archaic Ionic capital with


upright volutes,

an

member which was certo which we must assign the


The outline of the volutes

architectural

tainly antiquated at the period

execution of these sculptures.

and of the anthemion surmounting them

and there

clearly indicated,

is

upon the weathered surface of the

remain

still

stone traces of the engraved spiral lines which

The member

volutions of the helix.


precisely the
capital,

interior Troad,

and described

Archaeological Institute.^

the con-

thus represented

same shape and proportion

found by the writer upon the

mark

is

of

as the proto-Ionic

site of

Neandreia,

in the

in a separate publication of the

The

significance of this form

in

architectural history has been fully discussed in that connec-

In regard to

tion.
it

will

its

place in the composition of this relief

observe that the heraldic sphinxes of Assos

suffice to

paws upon a diminutive proto-Ionic stele, in the


same manner as the heraldic lions of Mykenai face a small
column having proto-Doric characteristics.
This arrangement was a common one in every age of
their

rest

Oriental

art,

and has by some historians been directly

ferred to an Asiatic origin.


too

numerous

to mention,

We

may

trace

it,

re-

examples

in

from Assyria and the highlands of

Asia Minor, through Phoenicia and Cyprus,


Peloponnesos.
at a period

It

to Attica and the


had certainly been adopted by the Greeks

long anterior to the building of the temple of

Assos, appearing not only above the gate of Mykenai, but in

many ornaments

of precious

metal found in the

ancient sepulchres of that citadel,^ and

among

Clarke, Proto-Ionic Capital,

fig. 2.

Schliemann, Myceitce, Nos.

175, 264-266, 274, 279, 480,

and

still

more

the remains

539.

arcHjEological institute.

176

unearthed at Menicli

in

uplifted

position

as

those

Sphinxes and

Attica.^

paws, in precisely the

same

occur

of Assos,

with

griffins

and

attitude

relative

upon the well-known

Francois vase,^ and are frequently to be observed upon vessels


of the early Corinthian style.

The

evidently

archaic Ionic capital with upright volutes

a favorite in the Troad and

its

neighborhood,^ whence the

known

majority of the

ex-

amples have been derived


finds

tion

still

coveries

may
ticed

The

another

among
at

in

be

no-

connection.

this

object in question

of

with upright ^j^


^f ^

Volutes.

is

terra-cotta

the

han-

amphora

(Fig.

vessel, apparently
Ionic Capital,

dis-

which

Assos,

appropriately

fragment
FiG. 42.

illustra-

minor

the

42).

The upright

scrolls are

Portion of the Handle of a large Jar found at Assos.

boldly modelled, the anthe-

mion, too small a feature to be fully indicated, being replaced

by a knob-like abacus.

The

spiral lines of the

volutes are

deeply ploughed out with some pointed instrument.

and abacus are

lightly

Scrolls

touched with a white pigment, the sur-

face of the red clay being elsewhere covered with a dull red

priming.
1

The

decorative effect of

Deiitsches archaologisches Institut.

ungsbericht von H. C. Lolling.

Roma,

the detail

Das Kufpelgrab

bet

is

striking

Meiiidi.

Ausgrab-

Athen, 18S0.

Mon.

Shortly after the discovery upon the site of Neandreia, Ionic capitals with

Ined., vol. iv.

upright volutes were

These

Das

liv., Iv.

found upon the coast of Lesbos, opposite the Troad.

capitals have not as yet

Puchstein (Otto),

1844-4S, pis.

been published, but they are referred

lonische Capitell.

Sichennndvierzigstes

Winckclmannsfeste der Archdologischen Gcsellschaft zu Berlin.


P-SS-

Programm
Berlin,

to

by

zttm
1887,

AT

INVESTIGATIONS

and, though slight in execution,

is

18S3.

in

the

77

work

of

Museum

at

evidently the

is

it

The fragment is now


numbered P. 41 21.

a skilled hand.

Boston, and

ASSOS,

In regard to the fragment of a sphinx from the western


front of the temple,
to the

Louvre

among

in 1835,

should be remarked that the litho-

it

in the

graph published

the reliefs removed from Assos

Monumenti

errs in omitting the up-

The engraving given by Clarac

lifted paw.i

makes good

this

defect, while the accompanying text supplies us with an all


too elaborate explanation of the significance of these mythical

animals

in this

As

connection.

types of intelligence and

perspicacity, Clarac holds that the sphinxes

whom

the story of Proteus, with

he would identify the marine

are to be taken as significant

and

of another relief,

monster

have reference to

the astuteness displayed by the hero Menelaos in over-

of

coming the wiles

of this prophetic old

man

In the

of the sea.

exposition of this view Clarac draws a parallel between the

Menelaos

feat of

in discovering

Oedipus

Helen

in

her Egyptian re-

solving the riddle of the The-

and
ban sphinx, even going so far as to suggest that the latter
episode may have been represented upon the walls of the
that of

treat,

temple

of

striking

We

Assos.

instance

of

in

scarcely

could

the

far-fetched

point

out a more

interpretations,

based

merely upon vague suppositions, which were in favor among


classical scholars as recently as the middle of the present
century.

The

researches of more

so clear a light

modern

sentations of the sphinx, and

1
iii.

Monumenti

Roma
Clarac,

archaeologists have thrown

upon the wide scope

Inediti pubblicati

deW

its

of the mythical repre-

general use in this form

Istituto di Correspondeftza archeologica, vol.

e Parigi, 1839-43, pi. 34.

Music de Sculpture antique

et

moderne.

partie.

12

Paris, 1841,

tome

ii.

seconde

ARCH^OLOGICAL INSTITUTE.

178

by ancient decorators,
Syrian, or Greek,

some

that we

Mesopotarhian, Egyptian,

need no longer thus grope

after

episodical connection in order to justify the appearance

among

of the sphinx

shown, a sphinx, in
fin,

whether

the Assos

reliefs.

later ages generally

As

the coat of

termed
of Assos.
was on

was the heraldic symbol

to-day be

will presently

transformed to a

It

arms, as

it

be

grif-

would

account that

this

these figures, in heraldic position and duplication, were sculp-

tured above the main entrances to that sanctuary which was

dedicated to the guardian deity of the town.


in its

Apart from

this,

independent significance, we need seek in this symbol

nothing beyond those characteristics of supernatural force,

wisdom, and ever-blooming youth which were ascribed by


popular belief to the sphinx as a combination of the bodily

forms of the strongest animal of the earth, the strongest bird


of the air,

and the

intelligent

head of a human being.

These

picturesque attributes of mysterious, almost demonic power,

seem

to

have been gradually connected with the sphinx

through the observation of a definite image, which had


arisen through the agglutinative

itself

methods of heraldry, rather

than to have been originally based upon any specific tradition.

Among

the ancients, no composite form was

more

From

widely known, none more frequently employed.


oldest Egyptian sphinx, the Colossos of Gizeh, a

work

that

of the

fourth dynasty, and from the winged Assyrian sphinxes of

the palace of Esarhaddon (681-668

b.

c), we

may

trace the

migration of the monster, and the development of the various


types of

its

representation in every part of the ancient world,

The

and

in every

the

human-headed and lion-bodied sphinx seems

been Egypt.

age of ancient history.

original

home
to

of

have

In the demonology of Mesopotamia the form

was never of more than secondary importance.


images always bear a foreign stamp, more or

The Assyrian
less distinct.

INVESTIGATIONS

AT

The winged type seems, however,

ASSOS,

1SS3.

79

have been derived from

to

Assyrian art by the Greek designers of the archaic period,

and
for

communication we are led

in the lack of direct

some intermediate stage

of development

to seek

and transference.

This can have taken place only upon those southeastern

where the

coasts of the Mediterranean

forms of both

hieratic

Egyptian and Mesopotamian art were blended in merely


decorative types, free from all fetters of religious symbolism.
It is

a well

known

fact that the favorite subjects of Hittite

and Phoenician decorators were constantly repeated

in

every

branch of Greek art-workmanship, during the archaic period,

and were retained long thereafter

in

distant Etruria.

The

hybrid art of Phoenicia, so widely disseminated through the

commerce

of

Tyre and Sidon,

is,

without doubt, responsible

Beyond

for the introduction of the sphinx to Greek culture.

the recognition of this fact,


possible for us to go

from the tenth

to

it

is

for

the present scarcely

the threads of artistic history which

the seventh

century before Christ con-

nected the civilization of Phoenicia with that of Asiatic and

European Hellas are so inextricably entangled

that

we cannot

attempt to assign to the races of Cyprus and southern Asia


Minor, more closely allied to the Greeks, their direct contributions to the transformation and introduction of this ancient
form.

We

may

transitional stage,

Euyuk a
link between the sphinxes of Nimroud ^
Spata^ and Etruscan Vulci,*
but we

recognize in

and those of archaic

the

sphinxes

of

are as yet far from possessing even so clear and succinct a


1

Perrot ( George ) Exploration Archeoloi^iqtie de la Galatie et de la Bithynie, d'rine


,

Partie de la Mysie, de la Cappadoce


pi.
2

et

du Pont;

execulee en 1861.

Paris, 1862-72,

65-67.

Layard (Austin Henry),

77^1?

Monuments of Nineveh,

vol.

i.

London,

1849, Pl- 448

Milchhbfer, in

the Mittheilungen des deutschen archdologischen Insiituts


Athen, 1879.
Micali (Giuseppe), Monumenti Inediti. Firenze, 1S44, pi. v.

vol. iv.
*

ARCHAEOLOGICAL INSTITUTE.

l8o

we

history of the artistic development of this form as

have,

for instance, in the case of the similar evolution of the Ionic


capital.
It is

only certain that in the sphinxes of Assos

we have

already to deal with a fully developed type, determined by

long

with

familiarity

the

rounded and bent forward

composite

The

form.

wings,

at the tips, are of a peculiar shape,

frequently met with in the oldest black-figured vase paint-

This conformation, most excellent

ings.

however much

at variance with all natural wings,

feathers or membranes,

In the

tendencies.

projecting

members

first

place

it

was necessary

to the panels

Thus

to

to

adapt the

upon which such figures


it

became

portant that the wings should rise but

which was

whether of

evidently the result of two distinct

is

were drawn or sculptured.


of the head,

in decorative effect,

little

particularly im-

above the crown

be made as prominent as possible.

In the second place, the conventionalization of the forms of


the living model for the purposes of decorative design
itself

particularly felt in such

made

irregular terminations as the

wing feathers. As the wings of the most ancient


Mesopotamian and Syrian sphinxes and grififins are not thus
rounded, this improvement is undoubtedly to be ascribed to
tips

of

that

Greek genius

left

its

for conventionalization

which everywhere

mark upon the material borrowed from the

The beginnings of the change


upon many painted vases and
style, as well as

to the period

in

East.

formation are to be traced

sherds of the early Rhodian

upon some few Phoenician works, referable


art of that country was influenced by

when the

Hellenic traditions and methods.

In various other details already referred

to,

both

reliefs

bear the stamp of that well trained yet somewhat conventional school of archaic design

which

in

remote and provincial

parts of the ancient world but shortly preceded the highest

AT

INVESTIGATIONS
development of sculpture

The very

it.

type and the du-

plication of the sphinxes adopted as the coat of

met with

are not infrequently

l8l

1SS3.

and the Peloponnesos, or

in Attica

was even contemporary with

ASSOS,

arms of Assos

in other parts of the ancient

The design on the Franbeen referred to, and, to name another

world during the archaic period.


cois vase has already

example, a tripod vase recently discovered at Tanagra, and

now

in

Museum

the

of Berlin,^ shows couchant sphinxes face

same

to face, in almost exactly the

position as those sculp-

tured above the entrances to the temple of Assos.

Troad

itself,

In the

such figures were naturally favorite subjects

and the modellers

with

the

Two

sherds of decorated vessels, showing the heads

painters of vases

wings of sphinxes

Greek remains

hand-made ware.

much

is

like those of the

similar manner,

in a

way

as the

fillet

of these heads, in particular,

Assos sphinxes

falls

upon our

the hair

and bound by a

encircling the forehead,

and

fragment of

Hissarlik,^ the first being a

The second

figurini.

have been found among the

of this design,

of

of

tainia,

behind the wings

Among

reliefs.

is

arranged

which, after
in the

same

the earlier dis-

coveries at Hissarlik^ was the figurine of a sphinx squatting

upon

haunches,

its

bling the

form

in

sphinx sculptured

the temple of Assos,

now

in

and

position

upon one
the

exactly

of the

Louvre.

To

resem-

metopes
these

of

may

be added seven further examples of squatting and couching


sphinxes, found by Calvert in various parts of the Troad.

Among

these the most

the Assos
1

reliefs

the

is

striking

to the

parallel

figures

of

sphinx painted on a sherd found

Loschcke (Georg), Dreifussvase aus Tanagra,

in

the Archdologische Zcitung,

Berlin, 1881.
2

Schliemann,

number 2379

Ilios, figs.

1432 and 1434.

^ ?iz\-\XKV(\'x\\x\,Trojanische

figurine

is

The former

of these

now

bears the

in the collection at Berlin.

Alterthumer.

now numbered 2433

in

Atlas.

Leipzig, 1874, no. 3362.

the collection at Berlin.

This

ARCH^OLOGICAL INSTITUTE.

I2

upon the ancient

Ophrynion

site of

(Fig. 43).

The

figure

here shown might almost pass as a direct copy of the couching sphinxes of the Assian reHefs, so close is the resem-

We

blance.

may

notice the

a convex curve, the

same

fillet-bound hair, falling in

same bordering

tainia, the

same wings

bent forward at the ends, and membered by a rib at the conventionalized juncture between flesh and feathers, the

doubly curved turn of the

Fig. 43.

same

and, in particular, the same

tail,

Couching Sphinx on Sherd from Ophrynion.

excessive emphasis of the rising curve of the belly, which

along

its

entire length

is

arched above the ground

ation of the natural appearance of

painting upon this sherd

is

in

makes

its

appearance

Another

the fleshy part of the wing,

in

between neck and wing feathers,


red,

The

mainly dark brown upon a light

red ground, the pigment being applied very thinly.


color

exagger-

recumbent animals.

similar to dragon's blood.

this

being of a dull deep

The

length of the

sherd

15 cm.

is

Near Erenkieui, between the

Alt-Trojanische Gr'dber

sites of

und Schddel,

Troy and Abydos.

Berlin, 1SS2.

Compare Virchow,

INVESTIGATIONS

AT

resemblance nearly as close

ASSOS,

is

183

18S3.

observable between the

squatting sphinx of one of the metopes of the Assos temple,


before referred

and a figurine from Aqkieui,

to,

Calvert as the site of Berytis (Fig. 44).


1 1

cm. high,

is

formed of a light red

clay,

intelligently modelled, belonging to the

from Hissarlik, before

figurine

ferred to,

by

hollowed within,

It is carefully and
same category as the

grayish black.

slip of

and primed with a

identified

This small image,

re-

and as another, somewhat

ruder, image of a sphinx found by

Calvert
lier's

at

Bounarbashi (Lecheva-

Almost the only

Troy).

ference between

this

form

sphinx and that shown

Assos metope
of

Oriental

dif-

of the

upon the

the high head-gear,

is

which

appearance,

is

seen upon the Berytis figurine, but

could

not find

frieze

of

place

sphinxes discovered

in

the low

in

The

the temple.

list

of

the Troad by

Calvert further includes four speci;

mens on black-figured sherds from


Akshi-Kieui, the

site

the later

of

Squatting Sphinx.
Figurine from Aqkieui.

Fig. 44.

(Hellenic) Thymbra.

Turning now from


arms

of

Assos as

of the chief

we

the

coat

of

among

appears

the ornamental sculptures

sanctuary of the town, to the heraldic symbol

stamped upon the


mintage,

it

coins,

from the

earliest to the latest

Assian

are at once struck by the fact that in the latter

case the image of a griffin

is

substituted for that of the sphinx.

In formation and position the bodies of the two mythical

animals are precisely

human being

is

alike,

exchanged

but upon the coins the head of a


for that of

an eagle.

ARCH^OLOGICAL INSTITUTE.

184
It is

change
writer

perhaps impossible to adduce a decisive reason for this

The most probable

of type.
is

able to advance

explanation which the

that the coins of Assos, circu-

is,

lating widely throughout the Troad, were, from the

printed with a griffin in order to distinguish


coins of another town of the province.

first,

im-

them from the

Upon

the coins of

Gergis, a sphinx forms the symbol of the obverse.

References

by ancient writers to matters of this kind are extremely rare,


but the fact of
is

its

appearance upon the Gergithian coinage

mentioned by Stephanos

specially

tiquity of Gergis,

of Byzantion.^

renowned as a stronghold

The

an-

of the Teucrians,^

and as the native place of one of the Sibyls, is beyond question, and the fact that it was one of the first towns of the

Troad

to

establish a

preserved in
in the

the

way

all

mint

is

attested by the archaic coins

Thus no

large collections.

difficulty stands

of the assumption that coins of Gergis, bearing

image

the

of

sphinx,

had been issued prior

emission of coins by Assos.

The

would naturally have precluded

its

to

the

adoption of this symbol


repetition

elsewhere in

Every precaution would, moreover, be taken


confusion of the coinages by the country

the Troad.

to prevent the

inasmuch as the values of the Gergithian pieces were


based upon a different standard from that adopted by the
folk,

Assians.^

Brandis

is

certainly justified in speaking of the griffin as

das eigentliche Stadtwappen of the

city,

but he

is

at fault in

assuming that the Assians did not issue coins before the
fourth century.

Coins of an archaic

ing from the middle of the

most

of the

fifth

series,

probably dat-

century, are to be found in

numismatical cabinets of Europe, under various

Stephanos of Byzantion, s. v. Tepyis.


Herodotos, V. 122, VII. 43.
8 Brandis (Johannes), Das Miinz-, Mass- und Grwichtswesen in Vorderasien,
Berlin, 1S66, pp. 310, 313.
bis auf Alexander den Grossen.
1
2

AT

INVESTIGATIONS

ASSOS,

They bear upon the reverse a

classifications.

1S5

1SS3.

lion's

head

an incuse square, and upon the obverse a couching

The

in

griffin.

discovery, during our excavations, of a coin of this type

with the inscription A^'2,, leaves no doubt as to the true


In the British Museum the rearrangement of
attribution.
the coins of Assos thus indicated has already been made.

The form

stamped upon the oldest known


shown in Figure 45 A. We here see a

of the griffin

coins of this series

is

creature precisely like the sphinx of the temple reliefs, ex-

cepting that the head

that of an eagle.

is

resemblance of the body

So

close

is

the

and modelling, that we

in attitude

are tempted to believe that the die-cutter intentionally fol-

lowed the design so familiar to Assians from


position

among

prominent

its

The

the sculptures of their temple.

of the fore legs

is

the same

action

the nearer one lying extended

on the ground, the one beyond being so uplifted as to rest


against the circular framework of the coin, in place of the

upright support.

The

curve

similarly salient.

the hip

is

belly rises in

the

same compressed

In like manner, the

tail

has the single turn, and the wings are rounded at the ends

and bent forward


edge by that
alized

fillet

at the tips,

being bordered along the fore

which may be regarded as a convention-

and elongated representative of the wing-bone.

short head, on the other hand,

type of Greek
unnatural.

that of the oldest

The
known

and

in part

Its features are peculiar,

griffin.

The mouth

disproportionately

is

long

is
;

widely opened, the lower jaw being


so long, indeed, that

if

closed

it

would project much beyond the hooked beak.


The protruding tongue curves upward; the outline of the jawbone
is

The one

prominently indicated.

ble appendage,

entirely foreign

resembling that of a hare.

ment and

vigilance.

It

ear visible

to bird

nature,

stands erect, as

a formida-

is

and rather
if

in excite-

further abnormal addition appears

ARCH^OLOGICAL INSTITUTE.

86
the

in

shape

of

knobbed

projection,

from

rising

centre of the skull, just above the large round eye.


is

upon

lion-headed

the

effectively

a reminiscence of the

perhaps

projection

proto-griffins

This

horn

single

Mesopotamia.^

of

the

It

enhances the energetic and defiant aspect of the

monster.
In

all

archaic

these details the head closely resembles

bronze discovered at Olympia,^ which

Figure 45

in

On comparing

B.

upon the Assian

tion

fine

sketched

and the representa-

this

certain

coin,

is

the

features

of the

latter

which might otherwise have appeared inorganic and inexThus, it is evident that the downplicable are made clear.

ward curve
could

of the lower jaw,

not possibly

into

fit

which, as before mentioned,

the

beak

closed,

if

is

derived

from an exaggeration of the similarly curved, but too short,


lower jaw of some head similar to the Olympian

Another point
ventionalization

upon the coin

of

the

same character

of the
in

is

bronze.

the strap-like con-

swollen cheek-pouch, which results

a distinct line, running

downwards from

1 Witness the figure upon the relief of a small temple near the palace of
Nimroud, referable to the age of Assurnasirpal (885-860 B.C.). Layard, Monuments of Nineveh, second series, London, 1853, pi. 5 and Discoveries in the
Ruins of Nineveh and Babylon, (Second Expedition,) London, 1853, p. 348 et seq.
Compare Perrot and Chipiez, Histoire de I' Art dans VAniiqiiite, Vol. IL Assyrie,
The horn appears upon similar images of archaic coins
Paris, 1883, p. 408.
of Asia Minor (ascribed to Miletos) referable to the seventh century [Numismatic Chronicle, new series, vol. xv., London, 1875, pi. viii.), and upon the lion's
;

head of the Kroisos mintage (Head, Barclay Vincent, A Guide to the Principal
Furtwangler (s. v. Gtyps, in
Gold and Silver Coins of the Ancients, pi. i.).
Roscher's Lexikon der Griechischen und Romischen Mythologie, Zehnte Lieferung,
Leipzig, 1886) is inclined to ascribe the adoption of this feature in the head of
the griffin to the Ionian Greeks of the Asiatic coast.
2

Ausgrabungen zu Olympia,

pi. xxvii.

This head

sentative of a distinctly pronounced and widely

is

the most perfect repre-

known

type.

Among

other

examples of the same form may be mentioned that shown by Salzmann (Auguste), Nicropole de Camiros, Paris, 1S75, pi. 43, and two in the Museum of
Berlin, Nos. 2935

and 1023.

INVESTIGATIONS
the

and

eye,

AT

ASSOS,

approaching

gradually

the

187

1SS3.

outline

of

the

jawbone.

Furtwangler

has shown this

peculiar form of griffin's

head to have been developed by Greek designers from the


Phoenician imafre of the

TT
mamtains
He

monster.

open

that

the

with

curved

mouth,

and

jecting tongue,

proa spe-

is

Greek

cifically

device,

referable to the seventh

century

before

The form

Christ.

of the griffin

which appears upon the


Fig.

known

earliest

Assos

thus belongs

A, Archaic Coin of Assos.

45.

coins of

(Enlarged two and a half diameters.

B,

to

a distinctly pronounced

Bronze Head of Griffin, found


AT Olympia.

The appearance

and widely known archaic type.


type, so

may
still

from

different

Greek world

at

that in general use

when

the period

throughout the

coinage was issued,

this

with good reason be considered as the retention of a


older stamp, and thus be held to indicate the employ-

ment

of

the griffin during

the sixth

century side by side

with the sphinx, and perhaps interchangeably with

Be

civic symbol.

representations

image

this as

of

griffin is that

which

is

of

known

Furtwangler

in vogue,

(Adolf),

geschichtliche BedeiUung.

it

is

certain

it,

as the

that in

all

subsequent issues the form of the

pean Greece as early as the

which continued

may,

exclusively adopted.

is

coins

it

kind upon the coins of Assos the

this

of the griffin

Upon Assian

of this

>ie

to

have been employed

fifth

in

Euro-

century before Christ, and

without material alteration, until

Bronzefunde aus

Berlin, 1880.

Olympia unJ deren

kunst-

ARCHAEOLOGICAL INSTITUTE.

i88

the latest ages of classic

art.

The

(Fig. 46.)

unnatural ap-

pendages, top-knot and upright ears, are here omitted, their


place being taken by a jagged comb.

The

that of a bird.

The wings

head has

entire

been so lengthened and flattened as more closely

resemble

to

are turned backward, and divided

The

along their entire length into distinct feathers.

often flung into the air in a double curve.

niscence of the older griffins and sphinxes


farther

not

fore

is

retained in the

is

which, though

leg,

by

supported

tail

curious remi-

stele,

or

pressed against the framework of


the coin,
air.

As

is lifted

aimlessly in the

the anatomist recognizes

the derivation of one species of

animal from another in the existence of rudimentary and useless


muscles, so also
of
FiG. 46.

decorative

may

the student

forms

the

trace

development of one type from

Coin of Assos.

(Enlarged two diameters.)

another

by the

appearance

of

features such as these, otherwise inexplicable.

With the exception


form

of a single variety, presently to

of the griffln

be re-

remained unaltered until the

ferred

to, this

latest

ages of Assian mintage, even appearing upon coins

which bear the heads

of Tiberius

and Claudius.

It is

ever to be remarked, that on such imperial coins the

which
ally

in earlier

ages invariably faces to the

turned to the

left, is

how-

griffin,

occasion-

right.

In the early years of the third century before Christ a


temporary fashion completely altered the stamp of the Assian
coinage, and led, as before mentioned, to the adoption of a
griffin of entirely different

appearance.

The head

of

Athene

upon the obverse, previously in profile, here turns to three-

/M'EST/GAIVOA'S
quarters face

recumbent

AT

1S9

ylSSOS, 1S83.

the griffin upon the reverse arises from his

position,

and strides upon

striding griffin of this

the

Phokaia, and

(Pig. 47.)

town

neighboring

of

of

all fours.

form appears upon contemporary coins

evidently

served as model to the diecutter of this Assian series.

single issue

to

have been made

only seems
of this

novel type, the mintage im-

mediately afterwards reverting to the accustomed im-

We

ages.

are enabled to

assign an approximate date

experiment through

to this

the

close

resemblance

Fig. 47.

of

Coin of Assos.

(Enlarged two and a third diameters.)

the

three-quarters

face

of

Athene, which appears upon the obverse, to the head upon


a coin of Antiochus
It is

I.

(b. c.

280-262)}

not possible to advance an entirely adequate explana-

tion of the reasons

which led

to the

temporary abandonment

of the time-hallowed Assian type, but

assume that

to

it

was due

we may be permitted

to those political motives

which

were of so great importance in this age of the fusion and


centralization

change

in

of

the

Greek

the civic symbol

The
states of Asia Minor.
may have been brought about

through agents of one of the earlier rulers of Pergamon, with


the intention of breaking down, in this as in

many

other

ways, the autonomous spirit of a town destined to be annexed


to that

kingdom.

2S3) the
1

Gardner

no. 12.

From

the

first

revolt of Philetairos

Pergamene dynasty had made


(Percy),

T/ie

its

Seleucid Kings of Syria.

(b. c.

influence felt on
London, 1878,

pi.

iv.

ARCHAEOLOGICAL INSTITUTE.

IQO

the coast of the Gulf of Adramyttion, rapidly extending

its

power throughout the Troad by reason of the services


rendered in repelling the continual inroads of the

The

it

Gauls.

Eumenes and Attalos must have

absolute dominion of

been preceded by many such attempts to counteract the

indi-

vidual and separatist spirit of towns which clung obstinately


to

the

rights of

local

independence.

this conservative feeling of the

Greek

In few matters was


citizen

more strongly

evinced than in the unaltering retention of the civic symbol

upon

coins, as indicated

An

Greek mintages.

by the archaic character

entire

change

of the

Assian symbol per-

haps being found impracticable, something might at

by altering the design, and assimilating

effected

many

of

least

be

to that of

it

a town upon the farthermost confines of the advancing state.

This town, Phokaia, had originally been a stronghold of the


lonians, and
at

Pergamon

it

was

in

to break

the interest of the cosmopolitan power

down

the distinctions of race, as well as

the tenacious hold of autonomous feeling.

highly remarkable combination of the forms of the two

varieties

of

griffin

which appear upon the Assian coinage,

namely, the couching and the upright or striding,


in

in

met with

is

the figures of a mosaic pavement of a building unearthed


the lower town. This building, which stood in close com-

munication with the Agora, was evidently employed

for

some

administrative purpose, and was therefore ornamented with

the civic symbol.

It

is

situated at a distance of only 23 m.

from the Bouleuterion, but on a much lower


ing provided to

it

level,

access be-

from the market-place by a subterranean

passage and staircases, designated JJ upon the general plan


of this quarter of the town, and shown in section upon the

drawing
edifice

into

of

the

western fagade of the Bouleuterion.

forms a quadrangle of 12.6 by 5.4m., and

two rooms

of unequal size, the larger

is

The

divided

and easternmost

AT ASS OS,

INVESTIGATIONS

I91

1883.

paved with the mosaic, has a clear space of 6.2, by


That the building was not a dwelling or shop, but
was devoted to the transaction of some public business, is
of which,

4.6 m.

evident from the entire lack of fireplaces, as well as from the

simple division of the plan, and the character of the interior

The masonry

decoration.

hidden from view by a revetment

rough, having been

plaster, painted with brilliant colors,


in detail

a subsequent chapter.

in

but exceedingly

substantial,

is

which

of

be described

will

The southern

wall,

to-

gether with the terraced street upon this side, has been entirely destroyed,

carried

away by the masses

of earth

and

debris which, falling from the Bouleuterion and other build-

ings upon a higher level, deeply buried the re-entering angle

and

of the rooms,
to a height

from

of

the masonry upon the north standing

left

doorway from the

1.2

to 0.5

m.

Traces of an entrance

street are, however,

to

still

be perceived

near the eastern end of the southern wall, from which side
the chambers must have been lighted through windows.
the hillside

of the structure

tire plan

rock, which

and

is

Museum

at Boston,
in

material are preserved in the

under the number S. 1155.

The

The

pave-

upon the quarried surface of

flooring consists of a thick substratum

upon which the mosaic pattern

rounded pebbles

The

of this

part, laid directly

the native rock.


of mortar,

be quarried from the native

to

used as a pigment by the Turks of

Specimens

ment was,

had

here of a bright red color, resembling ochre,

is

occasionally

Behram.

As

particularly steep at this point, almost the en-

is

of various colors,

imbedded

is

formed by

in a fine

cement.

lime of this cement was mixed with minute particles of

pounded

brick, for the purpose of diminishing the brightness

of the white

background

known

of preparation
1

visible

to

Vitruvius, II.

between the pebbles,

Roman
4. 3.

a mode

builders as opus signinmn?-

Columella,

I. 6.

12.

192

archjEOLogical institute.

:^?CT
-

-t^y

'1^
'

"^

-.rr-^
;--

1^

_-^^
-1^3

^
.

---:^-=;:;i^

Q
a
<:

Q
u
Q
<
u
S
w

'^''"Tf^^'^^p'''''

"^-;-sii^S

<

INVESTIGATIONS AT ASSOS,

1SS3.

plain border, one meter in width, formed of

193
mixed

light

gray and olive-green pebbles, surrounds the central pattern,

which occupies an oblong measuring 4.3 by 2.66 m. This


pattern itself, shown in Figure 48, consisted of two borders
of considerable width, separated

surrounding a central

The outermost
two

griffins

by narrow white bands, and

which measures 2.53 by 0.89 m.


cm. wide, contains upon each side

field

border, 51

face to face

the inner border, 22 cm. wide, a

wave ornament with large scrolls of many convolutions.


The central field was broken away in so great part that
was not possible

it

therein.

low,

It

is

to

make out

only certain

that

the

subject

brilliant

and white, were here employed

in

represented

colors, red, yel-

comparatively broad

surfaces on a dark green background.

The outermost and

broadest border

our

interest in the

subject of which has led to the consideration of the mosaic


in this

connection

is fairly

well preserved

throughout one

Almost everywhere the background

half of its extent.

of

dark olive-green was found to have held together better than


the figures, and in

several patches

where the pebbles had

been broken away the pattern could


impressions

side remaining

be traced by the

on the bed of mortar and the cement of the

left

Every

interstices.

still

detail of the

was thus

The monsters,

two

griffins

upon the longer

clearly recognizable.

alike in formation

and posture

of body, are

dissimilar in their heads, the one having the crest, elongated


skull,

beak, and wattles of a bird, the other a head resem-

bling that of a leopard or lioness, with widely open

mouth

and protruding tongue, and with curiously conventionalized


horns and beard.^

It

will

be recollected that similar pairs

Furtwangler

(s. v. Gryps, in Roscher) has pointed out that the Greeks derived
horned panther-headed or lion-headed t}'pe of the griffin from the Persians,
who in their turn had taken it from the Chaldean image of Tiamat, the enemy
1

this

of the gods.

Compare

the characterization of this monster given by DeHtzsch,


13

ARCH^OLOGICAL INSTITUTE.

194

of eagle-headed and

leopard-headed grifBns appear

among

the sculptured decorations of the temple of Miletos/ and


formed favorite subjects in later ages.^ The feathered wings,

turning slightly forward at the

are similar to those of

tips,

upon the more common coins

the griffins represented

of

Assos, while the farther fore legs are uplifted in the air like
those of the sphinxes sculptured upon the temple
tation

of the

this imi-

posture, while omitting the supporting stele,

same aimless and inorganic character which

resulting in the

has been noticed

in

the corresponding feature of the coins.

The hind legs, on the other hand, are erect and striding, like
those shown upon these coins, which, as before said, belong
to the first half of the third

abnormally long and


curve.

So

century before Christ.

thin, is flung into the air in

fine is the execution of the

The

tail,

a graceful

mosaic that even such

sufficient number of MesoWo


p.
potamian and Persian representations of the figure have been quoted by Furtwangler. This lion-headed and horned griffin undoubtedly found its way to the
northern coasts of the Aegean by way of Lycia, where it appears upon coins

lag das Paradies ?

assigned to the

first

Leipzig,

half of the

fifth

8S.

century

B.

c.

(Gardner, Percy,

TyJ'es of

and Fellows, Charles, Coitts of Aitcient


Reign of Alexander, London, 1855, P'- xi.), as well as upon

Creek Coins, Cambridge, 1883,

Lycia before the

iSSi,

pi.

iv.,

mintages of later date.


1 Chandler
(Richard), Ionian Antiquities, London, 1769, vol. iii. pi. vii.-.x.
Rayet et Thomas, Afilet et le Golfe Latmique, Fonilles et Explorations Archc-

ologiqties, Paris, 1877, pi. 17, 49-51.


2

of

Eagle-headed and leopard-headed griffins form pendants upon the reliefs


fine bronze helmets found in the Caserma dei Gladiatori at

one of the

Pompeii (Niccolini, Fausto e Felice, Le Case ed i Monumenti di Poinpeii,


Napoli, 1854, fasc. 19, tav. ii. 2, and ii. 7), upon the vase of Xenophantos (Gille and Stephani, Antiquites du BospJiore Cimmerien, St. Petersbourg,
1854, pi. 45, 46), upon a vase in the Museum of Berlin (Gerhard, Eduard,

Denkmdler des koniglichen Museums zu Berlin, Berlin,


and in a wall-painting (Zahn, Wilhelm, Die schdnsten Ornamente und merkmurdigsten Cemdlde atis Pompeji, Herkulaneum und
Stabicv, Berlin, 1828-56) and on a marble table of Pompeii (Niccolini, Case e
Monnmenti, fasc. vi.). They are likewise common upon Roman sarcophagi
von T. G. Wdcker,
compare Zoega (Georg), Abhandlungcn, herausgegeben
(Giuseppe Antonio), Monumenti Antichi,
Gcittingen, 1817, and Guattani
Roma, 1785, tav. iii.

Neuenvorbene antike
1836-40, vol.

iii.

no. 1791),

IXVESTIGATIONS

AT ASSOS,

1SS3.

In ar-

small details as the claws are distinctly indicated.


tistic

respects the design of the griffins

They

are

ment,

see the

drawn with a

muscles in

all

is

exceedingly good.

outstretched hind

leg,

and the swell of the

as

well as with a

appreciation of the principles of conventionaliasm.


skill is

move-

clear understanding of animal

parts of the body,

95

full

trained

likewise displayed in the adaptation of the figures to

the exacting methods of mosaic-work.

no attempt

Thus, while there

is

roundness to the limbs by shading, the

to give

concave curve of the wing and the distance of the farther


hind leg are ingeniously indicated by means of darker local

The

tints.

These
saic

outlines are everywhere vigorous

griffins

rank

among

and graceful.

the finest works of ancient mo-

known, and are without doubt

to

be assigned to that

period in which this branch of art attained

its

very highest

development.

Between the

griffins,

and

in

the corners of the border,

are six-pointed stars, of the shape familiar to

all

who have

amused themselves with

striking arcs from centre to centre

with a fixed radius.

is

modern carpenters

It

There

intrusted to their care.

the

human mind

with figures such as these tha^

delight to decorate architectural drawings

in

this

is

a singular fascination to

use of a pair of compasses,

so

accurately can the periphery of a circle be divided with six


strokes of

the

mechanical and
not,

opening by which

however, stand

design.

it

was generated.

inartistic character of these

The same

in

The

ornaments does

contradiction to the antique spirit of

six-pointed star appears not infrequently

upon Greek vases, and is to be seen, in monumental execution, upon the sacred buildings of Eleusis.

Enough remains of
the border to make
pair, precisely like

the griffins upon the shorter sides of


it

apparent

that

they were, pair by

those already described.

The dotted

line

ARCHAEOLOGICAL LXSTITUTE.

196
AA

(Fig. 48) indicates the longitudinal axis of the pattern,

and

will

fore legs

convey an idea of the proportions

and claws

of

one of the

griffins

be perceived upon the left-hand

of the whole.

beyond

The

this line will

The monster shown

side.

upon the right-hand side of the drawing has an eagle's


The corresponding feature of the griffin upon the
head.
has been almost entirely obliterated, yet from

left

lines

above

horns,

curved

undoubtedly representing the peculiar crooked

it,

we may

safely conclude

to

it

have been that of the

quadruped.

This altogether unnatural combination of the typical outstretched and uplifted fore legs with erect and striding hind
legs

that

is

to say, with conventional

forms which elsewhere

appear only during the early ages of Pergamene supremacy

leads to the supposition that the mosaic was executed during

the third century before Christ.

All other indications to be

gathered concerning the age of the pavement and the building which contained

chief

it

Apart from the

view.

argument

to the Agora,

are in entire

for this date

known

to

agreement with

this

design

the

artistic style of the

itself,

the close relation of the edifice

is

be a creation of the Pergamene period.

doubt that we have

There

is

of the

mosaic a further evidence of that period of the monu-

thus

little

mental renaissance of Assos.

It

in

was under

the fine figures

this

dynasty that

figured mosaics, which had previously been restricted to sa-

cred edifices, were generally extended to profane buildings.^

civic

the

first

flinty

hall

of the latter class to profit

by

this extension.

The

pebbles are deeply worn, and must have been trod-

den under
figured
1

such as this would naturally have been among

Although not

foot for generations.

pavement may even have remained

See upon

von IIer7}iann
there quoted.

this
Goell,

point

restored, the
in

sight until

Becker (Wilhelm Adolph), Charikles, neubearbeitet

Berlin, 1877, vol.

ii.

p.

143, with the ancient authorities

IXVESTIGAT/OXS

considerable period

demned

as

ASSOS,

1S83.

97

Moreover, the hall continued to be used

the Christian era.


for

AT

mosaic had been con-

after the

no longer serviceable,

for

it

was found

to

be

covered with other floorings of plain stucco, the removal of

which from the surface

of the design

was a work

of

some

difficulty.

The

fact that the conventional colors of the griffins, as well

as their forms, are indicated in the mosaic, permits us to con-

sider

them from another point


works of

in relation to ancient

when

it

is

borne

in

mind

of view, rarely to
art,

be obtained

and peculiarly interesting

that the conclusions thus derived

without doubt directly applicable to the polychromatic

are

treatment of similar subjects in sculpture.

grounds

of the patterns, as

also the

All the back-

pavement outside the

design, were formed, as has been said, of grayish green pebbles,

shading from a

tint

such as that shown by the inner

The darkest

side of an olive leaf to an almost perfect black.

of these pebbles were selected for the background, so that

the general effect of this was

much deeper and more

solid

than that of the outer border.

The

bodies of the griffins themselves were of round whitish

gray pebbles, of two distinct shades.

It is

apparent that an

attempt was thus made to indicate the spots which were


tributed to the griffins by ancient mythographers.

markings as similar

describes these

an animal
that the

still

common

in

to those

Asia Minor, and

it

at-

Pausanias

of a leopard,

may

well be

substitution of the head of a leopard for that of

an eagle was thus rendered more natural.


An indication
of such spots may be observed upon various other works
of ancient art.^
1

Pausanias, VIII.

2. 7.

For instance, upon the hic;hly remarkable Etruscan relief, published by


Braun (Emil), Pithtre Etriische Vulcenti, Annali dell" Institiito di Corrispoiidenza
Archeologica, Rome, 1859, and figured in the Monumenti Inediti, vol. vi., R6ma,
2

ARCHAEOLOGICAL INSTITUTE.

198

The wings

are of a very light bluish gray,

edged along the

front line of the bone with a bright yellow.

The same

yel-

low appears upon the beards, as well as upon the crest of


the eagle-headed and the horns of the lion-headed monster.

The beak

of the

former and the tongue of the latter are of a

Fragments of the mosaic showing these


as illustrating the method of adjoining the

brilliant red jasper.

well

colors, as

(No. in).

Museum

preserved in the

are

stones,

separate

In general,

work

of the figures is

finer

and smoother

it is

to

composed

in texture

at

Boston

be remarked that the mosaicof carefully selected stones,

than those of the background,

as well as brighter and clearer in color.

The

prevalent white tint of the bodies of the griffins

is

in

accordance with the descriptions of the traditional colors of


these monsters given by ancient writers.^

It

may

likewise be

observed, that in the rare specimens of ancient vase paintings,

where a number

pigments are employed on which

of

are represented, their bodies are white.

Upon

griffins

a vase pub-

by Jahn,2 the resemblance of the colors to those of


our mosaic is very close, the wings of the griffin being blue,
The same colors appear
while the rest of the body is white.
lished

upon the well known painted vase of Xenophantos,^ and upon


the fragments of an antique wooden sarcophagus,* as well as
1857-63, as well as in

Ans

de Fotdlles dans

Des Vergers
Maremmes

les

(A. Noel,) VEtnirie et les Etrjisques, dix


Toscanes, Paris, 1862, vol.

iii.

pi

27.

Com-

pare also the Pompeian helmet referred to in a preceding note.


1 Concerning the conceptions of the ancients in respect to the conventional
colors of griffins, compare Aelian, De Nat. Anim., IV. 27, quoting from Ktesias

and preserving the fragment of the hidica numbered xxvi.


Aelian

is

in

turn followed in the mediaeval tract

Philes, II., ed.

Pauw,

the colors of our mosaic,


2

No.

p. 15.
is

De Anim.

in Lion's edition.

Propr. of Manuel

different description, altogether

given by Ktesias, Jnd.,

at variance

with

xii.

Jahn (Otto), Ueber bemalte Vasen mit Gohischmtuk, Leipzig, 1S65,

pi. 15,

29.

Stephani, Antiquith

Published

in

the

dii

Bosphore Cimmerien,

same work,

pi.

84.

pis. 45. 46.

Further examples are supplied by

various painted sherds found in the same locality.

PI.

70 A.

INVESTIGATIONS
upon various

less

common

1SS3.

the

99

instances of a different usage are

witness the red and dark green plumage of two

inedited sphinxes in the Barbakion and Central

Athens.

important keramic specimens, too numer-

point, as

this

ASSOS,

Great stress should not, however, be laid

ous to mention.

upon

AT

It

is

possible that, as

to show,

would tend

Museum

of

the specimens referred to

the Asiatic and the European usages

differed in this particular.

The appearance

above the main entrance to

of the sphinx

the temple of Assos led to the remark that the figure of this

monster may have been employed interchangeably with that


This assumption
of the griffin as the symbol of the city.
derives further support from the appearance of the sphinx, in
its

characteristic crouching position,

The gem

discovered at Assos.
setting in which

it

upon an engraved

seal

broken from the

in question,

had originally been secured, was found

among the debris of the lower town by a peasant of Behram,


some years before the arrival of the American explorers. It
was purchased from the custom-house
by the present

writer,

official of

who, after mounting

it

the

little

port

in a gold ring,

copied from an antique of about the same age and character,

gave

it

Museum

to the

of Boston,

where

it is

preserved under

number S. 1020.
The stone is a carnelian,
a material much more highly
prized in ancient than in modern times.
It is worthy of note
the

that

Pliny

particularly

mentions Assos as one of the two

places in the Greek world

whence were derived supplies


under the name Sarda.

of

The

carnelian,

described by him

geological

researches of the expedition did not lead to the

discovery of any deposit of this stone in the vicinity of Assos.

and

it

appears probable that the carnelians

ancients as Assian were merely brought into


1

Pliny, Nat. Hist.,

XXXVII.

31, ed.

known to the
commerce from

Delph.

ARCH^OLOGICAL INSTITUTE.

200
this port,

and were found

where the formation

The gem

is

it

at

some place

in the interior

is

(Fig.

The

but slightly convex.


48'^) is

12

mm. on

and

of the tail,

carefully executed, being decidedly

attitude

is

which

is

it

belongs.

not uplifted,

thrown straight up into the

that of the sphinx sculptured

is

Fig. 48^

of the farther fore leg,

which

its

intaglio engrav-

superior to the average work of the class to which

With exception

Troad

not of a volcanic origin.

of an oval shape, measuring

longer axis, and


ing upon

is

air,

the

upon the epistyle

of

Heraldic Sphinx upon engraved Seal found at Assos.


(From the Impression.

the temple.

The

Enlarged

six diameters.)

wings, fully feathered, turn backwards like

those upon the later coins of Assos.


distinctly indicated.

The female

breasts are

Evident reminiscences of the archaic

image of the Assian sphinx are

to

be seen in various

such as the modelling of the haunch, and the

fillet

details,

binding

the head and falling upon the shoulders, as well as in the

general position.

There can be

was engraved with

little

definite reference

doubt that
to

this

the civic

gem

symbol

LXVESTICATIOXS A T ASSOS,

We

of Assos.
seal

cannot, indeed, adduce direct proof that the

was actually used

but there

nothing

is

root, as in the

city authorities,

the nature of the case to render this

identity of the

arms and

cients for coat of

by

in this significance

in

The

view unlikely.

201

18S3.

words employed by the an-

same

seal ring (or the use of the

Latin sigiatni and sigilhim) shows that the fun-

damental idea of such an image was heraldic

and there

can scarcely be a doubt that seals bearing the symbol or device of a boar were made use of not infrequently.
There

an

thus

is

agreement

entire

the mediccval usage


that

this

in

between

we have become possessed

Greek

We

of an

gem

are reminded by this

tyrant of Assos and Atarneus,

Persians

(b. c.

governors

into giving

that

actual

well

seal

of

and
be
the

of one of the most interest-

fell

into

When

Hermeias,

the hands of the

left in

up the

bearing the impression of

letters

its

seal

charge of those towns beguiled them

citadels

and garrisons, under the

belief

an amicable arrangement had been effected between

King Artaxerxes and


this ring of

inasmuch
adopted

upon

ancient

may

345), his enemies possessed themselves of his

and by sending

to the

it

city.

ing episodes in the history of the town.

ring,

the

and

respect,

as

for

their

Hermeias
the

their
to

rulers

former

ruler.^

We

may imagine

have borne the image of the sphinx,


of

such

small

states

commonly

purpose the symbol which they stamped

this

coins.

Diodoros, XVI.

It

52.

may be mentioned

Emperor Augustus likewise


image of a sphinx, two of which he had found
among the jewels of Atia (Pliny, Art/. Hist., XXXVII. 4, ed. Delph. ; Suetonius,
Octav., I.; Dion Cassias, LI. 3).
In the absence of the Emperor from Rome,
during the civil wars, his agents were authorized to use the duplicate to seal
official documents which had to be sent from the capital.
Pliny informs us that
it was a common jest among those who received such edicts that this sphinx
was ever the bearer of some enigma. On account of this mockery, Augustus
subsequently exchanged the sphinx upon his seal for another image.
employed

as a coincidence, that the

seal rings bearing the

ARCH^OLOGICAL INSTITUTE.

202

The interchangeable employment


first sight, is

we have the figures of the


Even a cursory examination

ring.

the sphinx and the

interchange.

to the temple,

civic hall

and

between

ages of antiquity,^

all

We

on the

of the seal

into the relations

intimate in

griffin,

will fully explain this

in

we have

While, on the one hand,

between the oldest coins of the town and the coat of

arms sculptured above the entrances


other,

griffins

appears at

it

thus to be traced in the Hellenistic as well as

in the archaic period.

parallel

and

of sphinxes

as the heraldic symbols of Assos, surprising as

may even

recognize

a constant tendency.

it

In mythological significance,

griffin

and sphinx were most

closely allied, and in that conventional artistic usage with

which we are

at present

concerned they were often regarded

The fundamental

as actually identical.

idea in both

The

parallelism between sphinx and griffin formed the subject of learned

investigations as early as the of time


vol.

iii.,

mon-

combining the characteristic features

sters because of their


1

that

is

of a supernatural, irresistible force, ascribed to these

Basileae, 1581, x. 62

length by Voss (Johann

xxiii.

Turnebus (Adrianus), Adversarioriim,


It has been treated at great

26; xxiv. 23.

Heinrich), Mythologische Briefe, Stuttgart, 1S27-34,

Anhang, Ueber den Ursprtmg der Greife, 2L.nd Theil ii. p. 189;
and more particularly by Stephani (Ludolf), Erkldrung einiger im Jahre 1863
im siidlichen Riissland gefundenen Gegenstdiide, Compte Rendu de r Academic de
Theil

i.

p.

305,

St. Petershoiirg, St.


_

Petersbourg, 1864,

The most important

P- 64.

materials for a consideration of the position and develop-

ment of this type in ancient literature and art have been collected, among others
by Ukert (Friedrich August), Geographic der Gricchen itnd Rdmer, Weimar,
1816-46; Welcker (Karl Gottlieb), Hekate tmd Eros, von Grcifcn gezogcn, in
his Alte Denkmdler erkldrt, Gottingen, 1849-64, vol. ii., and in another paper
entitled

Sarcophag im Museum

zii

Koln, Zeitschrift fiir Alterthumskunde des

Rheinlandes, Bonn, 1845; Baehr's note to Herodotus, 2d ed., Lipsiae, 1856-61,


vol. ii. ; Bruiiet's Recherches sur quelqiies Animaux fantastiqties, in the Revue
Archeologiqiie, vol.

ix.,

Paris, 1853

Brunn's Intorfio ad alcune

delta Sfinge, in the Bulletiiw delP Institttto,

Roma, 1853

rappresentauze

and, more particularly,

Stephani, in the works already quoted and in the volumes of the Compte Rendu
de

r Academic de

St.

Petersbourg for the years 1863 and 1867; Langbehn

FlUgelgestalten der ditesten griechischen Kunst,

as quoted above.

Munchen, 1881

(J.),

and Furtwangler

INVESTIGATIONS AT ASSOS,
most powerful animals

of the

are

among

and

griffin

the most

common

being identical

tion of the head,

it is

in

of earth

and

1SS3.

ity

Hence both

air.

Sphinx

apotropaic symbols.^

formation with the sole excep-

not difficult to account for their similar-

The

of attributes and exchange of functions.

head of a leopard with that of an eagle,

of the

203

of our mosaic, indicates the readiness with

alternation

in the griffins

which the human

head might be exchanged for that of the bird.

As

the result of this

we may frequently

literature a failure clearly to distinguish

we constantly see sphinx

sters, while in ancient decorative art

and

griffin

notice in ancient

between these mon-

employed as the most natural pendants, and often

used interchangeably.

Thus

it

ences to
dKpajr]<i

is

Kvwv^

that applied
7t-pvTavi<i

and

worthy of note that

the griffin

by the same author

word.*^

fusion

is, it is

selves

but

one

of the earliest referit

^varo[xo<i

Zr]vo<i

a designation which closely corresponds with

Kvwv^

The Romans

to the sphinx, hva'qfiepiav

certainly conceived sphinx

be identical, designating both by one and the

griffin to

same

in

Aischylos terms

So

direct literary proof of this particular con-

true,

wanting

in the case of the

noteworthy parallel

is

Greeks themby their

presented

identification of the griffin with the hippalektryon.^

We

reach these same conclusions more directly, and with

For much that concerns the prophylactic signifiance of sphinx and griffin,
Die Lauersforter Phalerae, Bonn, i860. Sphinxes are very
frequently represented upon apotropaic vases.
1

see Jahn (Otto),

Aischylos, Prom., 803.

2
3

Aischylos, Frag., No. 232.

word

the
*

XX.

Sophokles {Ocd. R., 391) applies

" Piceis," " Pices," or " Phices," [Festus, ed. Miiller,


p. 206
2,

Nonius Marcellus,

De

Propr. Sertn.,

to the important passage of Plautus, AtiL, IV.

the
*

to the

sphinx

pa}p({)5di kvwv.

p.

Isidorus, Orig.,

152. 7, Leipzig ed.,

who

refers

from the Boeotian form of


as used, for instance, by Hesiod, Theog., 326.
3.

i),

word sphinx, </)if,


This is done by Photios, Lex., and by Hesychios, both

s.

v.

'iTTTaXeKTpuwv.

ARCH^OLOGICAL INSTITUTE.

204

greater certainty, in surveying the wide field of ancient deco-

number

rative art, a small

which
off

of characteristic

will suffice for illustration.

mahgn

led to the

examples from

The potency

and

influences, attributed alike to sphinx

employment

of both

forms

tions of arm-chairs, couches, lamps,

warding

in

griffin,

in the sculptured decora-

and other furnishings

of

the dwelling, as well as upon the helmets and breastplates of

the warrior, and

it

was without doubt through

significance of an occult protection, quite as

this

much

common

as through

the resemblance of form, that the identity of character was


established.

commonest functions, as the guarmonuments or of sacred edifices, the griffin

In one of

dian of funeral

formed a pendant
gether.

It

its

the sphinx, or even replaced

to

be recollected

will

that

acroterion of the temple of Assos

which

of these

the

in

we were

it

alto-

case of the

doubt as to

in

monsters the fragment of a paw was to be

ascribed.

In apotropaic vases, of the earliest as well as of the most


recent styles, the griffin

with

the

of both

Apart from Oriental examples

sphinx.

parallelism^ well

monsters

constantly found in combination

is

known Greek examples

of the

of

this

appearance

compositions are afforded by the

in figure

Francois vase, before instanced, by a vase of the Ermitage,


published by Micali,^ and by one from the Castellani collection.^

and

It

must have been

griffins stood together in the palace of the

Skyles at Olbia,*
1

as such pendants that sphinxes

if,

indeed,

we may

Layard, Monuments of Nineveh, vol.

(Alessandro Palma

di),

Salatninia,

Baptiste Feli.x), Recherches sur

le

Orient et en Occident, Paris, 1867,


Micali, Moniitnenti, tav.

pi.

i.

89; vol.

London, 1SS4, 2d
et al.

ii.

pi.

69.

For an Etruscan example, see

xliii.

Micali, Afonumenti, tav.

Frohner (Wilhelm),

xl.

Collection Castellani,

Cesnola

Lajard (Jean
Mysteres de Mithra en

ed., fig. 115.

Ciilte piibliqtie et les


fig. 58,

Scythian King

not conclude, from

Herodotos, IV.

Rome,

1884,

No.

368.

78.

INVESTIGATIONS AT ASSOS,
the frequency of

grififins

205

1SS3.

and the entire lack

of

sphinxes

the remains of this part of the ancient world, that

among

Herodotos himself affords us an example of that confusion


between the two forms which appears to be frequent among
ancient writers.^
the sphinx, with which

it was so readily
means originated amongst the Greeks,
This fact is substantiated not only by the constant appearance of griffins upon Oriental monuments of the highest

The

like

griffin,

confounded, by no

antiquity,^ but also


1

Compare upon

by the traditions of an Eastern derivation


passage given by Rawlinson in his

this point the note to this

translation of Herodotos, London, 1S75,

^''- iii-,

and the authorities

in

regard to

the discovery of griffins in Scythia there cited.


2

An

outline history of the

employment

recognized from the facts already adduced.


like

the sphinx, undoubtedly had

its

of this form in ancient art

To sum up

concisely.

origin in the agglutinative

may be

The

griffin,

methods

of

mythology and heraldry. The earliest formal combination of the kind known
to the writer is that Chaldean image of the winged lion with bird's talons upon
In Assyrian art
its hind legs which Assyriologists identify as the enemy Tiamat.
we see this figure assume the eagle's head (Lajard, Cu/te de Mithra, figs. 54 B,
56, 57;

Menant (Joachim), Les

vol.

ii.

11),

and

fig.

finally

PArt dans

adopt the characteristic form of the

of a lion (Layard,

Monuments of Nineveh,

Ruins of Nineveh and Babylon,

An

Pierres Gravies de la Haute Asie, Paris, 1883-86,

Perrot and Chipiez, Histoire de


vol.

i.

VAntiqidte, vol.

griffin

pis. 8, 43,

46

ii.

fig.

with the hind legs


;

Discoveries in the

p. 200, et al.).

adequate investigation into the further history of the type would here

lead us too far


griffins, closely

afield.

Attention should, however, be called to the fact that

approaching

in

form those depicted by the Greeks of the archaic

period, and differing in certain important respects from the

Tiamat type of
Mesopotamia, appear upon the most ancient engraved cylinders of the Hittites.
Lajard, Culte de Mithra, fig. 58
Wright (William), The Empire of the Hittites,
London, 1884, pi. i. ; Seal from Marash in the Museum of Berlin, No. 7894.
Perhaps the next stage in the further migration is presented by the sitting
griffin so frequently met with upon Cyprian seals, such as that referred to in a
former note. In general it may be remarked that the griffin quietly posed as a
guardian, and not in attack as a beast of prey, forms the Syrian, as contrasted with
the Mesopotamian type.
It is this guardian which appears upon the seals of the
very earliest Greek period, such as the " island stones " in Copenhagen, Breslau,
and the British Museum, published in the Archdologische Zeitung, Berlin, 1883,
pi. 16
and upon the Boeotian tablet mentioned by Milchhcifer (Arthur), Die
;

Anfdnge der Kunst

in Griechoiland^ Leipzig, 1883, p. 48,

and now

in the

Museum

ARCH^OLOGICAL INSTITUTE.

206

dwelt upon by almost

who make mention


archaeologist

is

all

those classic writers, before quoted,

The

of the monster.

hence rather

to

task of the Greek

follow the changes which

and perfected the primitive Oriental

altered

seek for direct explanations of


signification.

logical

type, than

origin, or original

its

Taken from

this

to

mytho-

point of view, the

great diversity of the sphinxes and griffins of Assos renders

them

of peculiar interest.

attributes of the chief deities of ancient towns were

The

commonly adopted
Athenian

as civic

symbols

Pallas, the steeds of

for

example, the owl of

Delian Apollo, the peacock of

Samian Hera, the cock of Epidaurian Asklepios, the stag of


Ephesian Artemis, the goat of Ainian Hermes, and many
This usage leads us to inquire whether the sphinx or
others.
griffin

may

not have been

upon

this Asiatic coast, at least,

where these monsters were early received from the East,


and possibly connected with an orientalized
with the protecting deity of Assos.
fact,

Some

cult

associated

indications do, in

point to the existence of such an association.

Chief,

most familiar among these is the general adoption


of sphinx and griffin as symbols upon the helmet crest of
Athena, as in the chryselephantine statue of Pheidias.^ a work

as also

such details were most carefully considered.


Some other points, capable of supporting this argument, have
been referred to by Stephani.^ But after examination of all

which

in

all

instances of such association,

we

are forced to the conclusion

that the available materials do not suffice for a definite attri-

bution of either sphinx or griffin to Athena, or indeed any


The small griffins of beaten sheet gold found among the
of Berlin, No. 7548.
most ancient remains of Mykenai (Schliemann, Mycenae, No. 261) are more
Oriental in character.
1

Pausanias,

I.

24. 5.

Stephani (Ludolf), Der aiisriihende Herakles, pp. 147, 182, and the other
publications of this author already quoted.
-

AT

INVESTIGATIONS
one Greek

deity.

that existing

207

188S.

Perhaps the closest relation recognizable

doubtless

in great

attributes of invincible force


sters

ASSOS,

between
The

and Phoenician Herakles.

is

measure by virtue of their

mon-

these Oriental

exploits of this hero, as

be borne in mind, formed the subjects of the sculptures

will

of our temple.

Whether

it

was

in

any wise through

this

intimate connection of Herakles with the official worship of

Assos, natural enough in an Aeolian colony, that sphinx and


griffin

were adopted as

civic symbols, can, in the present state

more than a conjecture.


Here we may terminate our consideration of those temple

of our information, be no

sculptures which were discovered during the excavations of

the second and third years, and of those other works of art

which by reason of their subjects are connected with them.

The

known

following observations upon the reliefs previously

have the sole purpose of correcting and supplementing the


accounts hitherto published.

Of the

epistyle blocks

and metopes seen upon the

the temple by earlier investigators, there

is

been impossible to discover, or at

events

Prokesch von Osten,

in the

reliefs

" ein sitzender

1826,^

which were

the surface of the earth.

Amor, der

Of
die

is

nature
The

doubt whether a
actually

relief

disappeared

identify.
visit to

described as

auf den Bogen


this

stiitzt

was found either

We

by the French or by the American explorers.


in

are thus

showing a subject

during

has

be seen upon

at that date to

Hand

it

enumerates eleven frag-

these the eighth

Nothing exactly corresponding with

left

to

two descriptions of his

the site of Assos in the year

ments of the

all

site of

one which

the

nine years

of

this

which

was given in the Anzeigeblatt of the Wiener


Wien, 1832, under the title, " Mittheilungen
aus Kleinasien von Oberstlieutenant von Prokesch Osten
I. Reise von Smyrna
durch Mytilene nach Alexandria-Troas und Assos," July, 1826 and subsequently
in his Denkwiirdigkeiten und Erinnerungen aus dem Orient, before quoted.
1

earlier of these accounts

yahrbiich der Literatur, vol.

Iviii.,

ARCH^OLOGICAL INSTITUTE.

208

elapsed between the visit of Prokesch and that of Texier, or

whether Prokesch may have seen a portion of the


Herakles and Pholos (Fig.

discovered

37),

relief of

no great depth

at

beneath the surface during the American excavations,

and

have entirely mistaken the position of the body and

failed

to discriminate

between the two personages.

of Eros should have been represented

the temple

is,

in itself,

among

bowman

kesch to have been the Aeolic hero.


that

It

some carved blocks belonging

been removed from the


ent generation.

the sculptures of

exceedingly improbable, and

perhaps warranted in assuming the

known

That a figure

site

Indeed,

are

seen by Pro-

moreover, well

to the

temple have

by Turkish builders

of the pres-

is

it

is,

we

much

still

during the

first

surprising that so

remains.

The one complete metope


year was referred to
ing a

man

relief discovered

Preliminary Report as represent-

in the

pursuing a woman.

block has, however,

made

Further examination of the

apparent that the figure upon the

it

supposed to be that of a female,

left,

amounted
known, naked female

to a proof that

in the

it

figures

was

is

also

This

entirely nude.

male

for,

as

is

well

were not depicted by the Greeks

age to which these sculptures belong.

Even

as late

as the time of Praxiteles the nudity of the Knidian

Venus

required to be explained and justified by the suggestion of

the bath.

Close scrutiny of the stone showed that the figure

of the pursued, like that of the pursuer, had been originally

provided with male organs, which must have been obliterated


at

some period subsequent

to the displacement of the frieze.

Several of the other reliefs have suffered similar mutilation.


In order to complete our view of these sculptures,

we must

give our attention to two of the reliefs from the temple


Assos, removed from the surface of the ground by the

now
of

French

in 1835,

and now preserved

in the

Louvre.

The

sub-

jects of these remarkable works, in the writer's opinion, have

AT

LW'ESTIGATIOXS

ASSOS,

not hitherto been correctly explained.


their significance

is

account, but because


cation

of

of importance, not

determination of

merely on

its

own

has a bearing on the age and dedi-

it

temple, and

the

209

1883.

on the

relations

artistic

of

its

sculptured decorations.

The

chief of these reliefs,

upon the longest known epistyle

block of the temple, represents a marine monster, a kind

merman with human body and

of

seized from behind

space beyond and above the

human

retreating

figures,

arms, as

if

affrighted

at

merman and hero

thirty degrees,

same

level,

size of the

and as

with

the

filled

outstretched

As

the struggle.

are inclined
all

has been

The mon-

to

heads

panel, the

the

half

being

tail

who

tail,

(Fig. 49.)

occupies considerably more than

ster

the

fishy

by a naked hero.

in

with six

and

uplifted

the trunks of

an angle of about
to

rise

precisely the

the six upright figures are of less than half the


hero.

This want of scale results from the de-

signer having adhered to the conventional principle


as isocephalism

lowed

in

many

fol-

archaic works, and often recognizable, though

skilfully disguised,

this case

known

a method of composition regardlessly

even

in reliefs of the perfected style.

the naive violation of relative proportions

is

In

not

without advantage; for the figures of the chief actors in the

scene are thus rendered prominent in the same striking fashion as are the heroes of

and

reliefs of

those monumental wall-paintings

Egyptian and Assyrian

king strides victorious through hosts of

in

art,

pygmy

which a giant
assailants,

and

warriors outtop the fortification towers which they defend.

The
that

it

surface
is

of

the

stone

is

chipped and

so

corroded

scarcely possible to determine the sex of the six

retreating figures.

Texier^ and Clarac^ describe them as

Texier, Description de VAsie Mineure, vol.

Clarac, Musee de Sculpture, vol.

14

ii.

ii.

^
1

'[^y^^^

Pi

i'.;|^V

/'7

fli/i.

u
s
H
o

P!^

INVESTIGATIONS
female

De

Witte,

former view, and

it

as

male.^

is

at least

AT ASSOS,
The

1883.

writer

21

inclines

the

to

certain that the ensfravino-s

published in the Momimoiti^ are incorrect in showing these

As

bodies as nude.

indicating the terror inspired by the

struggle, these fugitives effectively emphasize the

main

ac-

In decorative respects, they contrast strikingly with

tion.

the slanting trunks of the combatants, being erect, or slightly


inclined from the group, the

arms of the

first five

a contrary direction, while those of the

in

the

are

relief,

postures

is

held vertically

aloft.

The

last,

outstretched

terminating

regularity of the

almost that of a conventional ornament

even

the turn of the heads, towards or from the dreaded spectacle,


in

is

unvaried alternation.^

The attacking hero, though entirely naked, bears upon his


back a quiver, seemingly rather as an attribute than as part of
his equipment, and is thus sufficiently designated as Herakles.
Such was

not, however, the identification of this figure given

by the

earliest editors of the relief.


Both Texier and Clarac
conceive the wrestler to be King Menelaos, and the monster

to be

one of the shapes

reaved husband
the Odyssey.*

is

of Proteus, the

Egyptian,

whom

the be-

constraining to prophesy, as recounted in

In conformity with this view, Texier supposes

the six fugitives (whose figures are, as he declares by

way

of support for the identification, tout a fait Egyptiennes) to

represent Helen and her maids, at that time domiciled with


Proteus, after having been driven to

Egypt by contrary winds,

while on the voyage from Sparta to Troy.^

Clarac, on the

other hand, conjectures that the six subordinate personages


1

De

Momimenti,

It is

Witte, Annali, 1842.


vol.

iii.,

1839-43.

a further error of the engraving in the Monumenti, that the fourth

ure from the combatants


*

Odyssey, IV. 435-460.

Herodotos,

is

II. 112, 118.

shown lookmg forwards.

fig-

ARCH^OLOGICAL INSTITUTE.

212

may be taken for the daughters of the Old Man of the Sea,
"among whom we may recognize the divine Eidothea" (the
only daughter of Proteus known to Homer, by the way)
;

or,

again, six of the Nereids

" constellations

Hyades,

or the

one prefer, the Pleiades,

or, if

which, by reason of the

may

changes produced by them in the seasons of the year,

be brought into relations with Proteus, the variable god, to


whom they may have been compared, and with whom they

may have been

associated."

Yet, perplexed by the fact that Menelaos

known

not

is

fame as a huntsman, while a variety of animals, lions,


boars, bulls, and the like, are figured among the sculptures
to

of

the temple, Clarac

an alternative an entirely

as

offers

different identification of the hero,

Aristaios,

son

imitation of

of

who he suggests may be

Apollo and Kyrene.

Virgil,i

the Homeric episode, has described

in

evident

this bucolic

divinity as questioning the prophetic Proteus concerning bee

and Aristaios thus might be represented

culture,

in the strug-

gle with the marine monster quite as naturally as Menelaos,

Clarac calls attention to the

quiver

is

that the attribute of the

fact,

decidedly more in keeping with the functions of

Aristaios than with those of the Lacedaemonian king

the six fugitives of the

may

relief

remain,

as

while

before,

the

daughters of Proteus, the Nereids, the Pleiades, or the Hyades,


or they

may be

taken as representatives of the Muses, or

whom

of the Seasons, from

Aristaios had received his train-

ing in the arts of husbandry.

Every indication afforded by

the decorations of the temple

is

with this explanation.

Thus, the

ravaged the country around

connected with the


^

lion is that beast

Mount

Kyrene, the mother of Aristaios


ans, naturally

readily brought into line

Pelion,

which

and attacked

the Centaurs are Thessali-

nymph and

Georgics, IV. 3S7 et seq.

her son, inas-

INVESTIGA TIONS A T ASS OS,


much

1SIJ3.

as Cheiron had himself instructed the

boy

the pairs

of bulls arc those offered by Aristaios in sacrifice before enter-

ing upon the struggle with Proteus

and, finally, the pig and

the stag are peaceable animals, significant of the agricultural

renown

who

of this mortal,

attained to the dignity of a god

through the benefits bestowed by him upon mankind.


exegesis such as this

is

An

too perfect and too characteristic to

be passed without mention.

On

the other hand, the sufficiently obvious fact that the

hero represented upon our relief

was

recognized by

fully

De

none other than Herakles

is

Witte, one of the

earliest editors

of the Assos sculptures, and, as compared with Texier and


Clarac, a trained archasologist.

De

attacks was, in
like that other

Old

The monster whom

Witte's view, Nereus,^


]\Ian of

the hero

being gifted,

the Sea, Proteus, with powers of

prophecy and of transforming

shape to elude the grasp

his

known, fabled by
the later Greeks to have sought from Nereus advice concerning the whereabouts of the Hesperides and golden apples,
of mortal hands.

and for

this

Herakles was, as

purpose to have

asleep, holding

is

well

upon him while he was

fallen

him firmly during

his

various

This identification of the monster as Nereus, correct-

tions.2

ing as

it

did the

most obvious error

handbooks, and has been reiterated


such

The
1

as,

among

its

way

in recent years

into

many

by author-

and Murray.*

others, Lenormant'^

fact was, however, soon recognized that the scene thus

In respect to this marine opponent o Herakles,

Nereus

Menelaos-Proteus

of the

version, was generally accepted, found

ities

transforma-

leaves the choice open between

first, still

failure to decide

between the two candidates

is

De

Witte, though naming

him and Triton.

The same

noticeable in the description of

the relief given by Guigniault.


-

Apollodoros,

Lenormant

II. 5. 11.

(Fran9ois),

Scholiast to Apollonios of Rhodes, Arg., IV. 1396.


Jntailles

Archaiques de

I'Archipel

Grec,

Revue

Archeologique, vol. xxviii., Paris, 1874, pi. 12.


*
Murray (Alexander Stuart), History of Greek Sculpture, London, iSSo-83.

ARCH^OLOGICAL INSTITUTE.

14

appearing
defined

among

the sculptures of Assos belongs to a well

and Gerhard,^

category of archaic representations,

republishing in the year 1843 a vase painting which resembles


the Assos relief in several characteristic features, satisfactorily

proved that the monster suffering under the rough embrace


of Herakles

His argument was

not Nereus, but Triton.

is

based upon the observation, that in no case does the group


display indications of that illusory change of shape by which

Nereus
tis

resisted Herakles, as Proteus did Menelaos,

To

did Peleus,

this

to

is

be added, that upon three of the

vases depicting this struggle the

name

of the

by an

monster

appears

inscription,

in

Gerhard, Anserlesene

''

Dass

De

ii.

pi.

cxi.

of

Hence

Triton

ein ivirklicher

Griechische Vasenbilder, vol.

had, in the previous publications of

given

several

these representations as a spectator of the combat.^

Gerhard justly concludes:

is

Moreover, Nereus

by accompanying inscriptions as Triton.^


himself, identified

and The-

in

This vase

Witte, Description des Antiqiiitcs

et

Ob-

d'Art qui composent le Cabinet de Feu M. le Chevalier E. Diiraud, Paris,


1S36, No. 302, and Dubois (Leon Jean Joseph), Description des Antiques faisant Partie des Collections de M. le Co7nte de Pourtales-Gorgier, Paris, 1S41,
No. 196, passed as representing Herakles and Nereus. Gerhard, in his Berlin^s

jets

Antike

Bildiverke

monster

upon

it.

bilder des

Berlin,

besc/trieben,

of the vase

No. 697

"

Nereus,"

1836,

had

like

in

manner called the


"Tritonnos"

in spite of the inscription

Compare also the same author's


Museums zu Berlin, Berlin, 1843,

Etruskische
pi.

xv.

und Kcimpanische

Vasen-

5, 6.

Museum, referred to in the precedmore recently described by Furtwangler (Adolf), Beschreibung der Vasensammlung im Antiquaritim, Berlin, 1885, No. 1906.
For
2

One

of these

is

the vase in the Berlin

ing note, which has been

the second, see Brondsted (Peter Oluf), Description of thirty-two Ancient Greek
lately foicnd in Excavations made at Vtdci, in the Roman Territory,
Mr. Campanari, London, 1S32, No. 7; also in the Archdologische Zeitung, 1856,
For the third, see De Witte, Description d'une Collection de Vases Peints
24S.

Painted Vases,
by
p.
et

Bronzes Antiques proz'enant des Fouilles de I'Etrurie, Paris, 1837, No. 84.
3 As, for instance, on the vase last referred to in the preceding note.
That

on another vase, published by Dubois, A'otice d'une Collection de Vases Antiques,


Paris, 1843-48, the name Nereus does not appertain to the monster, but to one
of the lookers on, has been pointed out by Jahn (Otto), Berichte der K. Sacksischen

Gesellschaft der

r'esuml of the

argument

IVissenschaften,
for this

Leipzig,

change of names

1854,
is

p.

173.

An

excellent

given by Roulez (Joseph

AT ASSOS,

INVESTIGATIONS

215

1883.

dicsan geqjidlten Meergott gevteint sei Icidet soinit durcJimis


Ziveifeiy

keiiicn

Reasons
works

for

questioning the appearance, upon such archaic

of art, of

Nereus engaged

in a struggle

with Herakles,

are likewise to be derived from purely literary sources.

Long

before the publication of Gerhard's argument, VVelckcr^ had

advanced the theory that the legend

of

Nereus prophesying

Herakles was merely an imitation of the Homeric story of

to

Emmanuel

Ghislain),

Royale des Sciences


^

Welcker

Museum

(F.

Lutte d'Hercule

et

de

Triton,

Bulletin de

r Acadcviie

et Belles Lettres de Bruxelles, vol. xi., Bruxclles, 1S44.

G), Die zwblf Kdmpfe

fiir Philologie,

Geschichte

und

des Herakles bey Pisander, Rheinisches

Griechische Philosophie, Bonn, 1833, re-

printed in the second edition of his Akademische JCtnistmuseum zu Bonn, 1841,

Welcker apparently overin his Kleine Schriften, vol. i., Bonn, 1844-50.
looked the fact that the scholiast to Apollonios of Rhodes [Argon., IV. 1396), in
repeating the tale of Nereus prophesjang to Herakles, refers to Pherekydes as
and

his authority for this version of the story.

Pherekydes the logographer (not to

be confounded with the better known philosopher of the same name) lived during the first half of the fifth century before Christ, and it is hence evident that
this episode of the

Herakles legend

is

older, at all events, than that large class

of literary embellishments and duplications for which


plagiarists of the Hellenistic age.

On

the other hand,

we are indebted to the


we learn from Athenaios

(XI. 38, p. 469 d), that Panyasis the poet, a contemporary of Pherekydes, relates
that Herakles received the sun-bowl for the voyage to Erytheia from Nereus.

This

may

well have been the original version of the tale, which but loosely con-

nected the adventurous hero with the Old

remarks, that

"

Man

of the Sea.

Athenaios himself

perhaps, as Herakles was fond of large cups, the poets and

torians, jesting because of the great size of this one,

fable of his having

gone

to sea in a cup."

may have

his-

invented the

This vessel was naturally reputed

to

have been given to Herakles 'by Okeanos, as stated by Peisander (Athenaios,


XI. 38),
a much older and better authority than either Pherekydes or Panyasis,

and
nos.

in this

From

business Nereus

is

obviously nothing more than a deputy for Okea-

this genesis of the tale of

Herakles and Nereus

it is

plain that the

connection between the two was neither sufiiciently primitive nor sufficiently

upon such
must have been.
In regard to the marine monster itself, Welcker, though rejecting all connection between this combat and the expedition to the Hesperides, yet fails to make
any advance towards the true solution of the problem. He still follows the identification of the merman shown upon the ancient vase-paintings as Nereus, contenting himself with the generalizing remark that the labors of Herakles led him
to subdue the monsters of the sea as well as those of the land.
close to serve as the basis for a scene so popular as that represented

numerous works of archaic

art

ARCH^OLOGICAL INSTITUTE.

2i6

Menelaos and Proteus, invented as an embellishment

of the

exploits of Herakles.

Gerhard's identification of the monster


so firmly established, that Klein

is

now

regarded as

has based upon

it

the bold

yet plausible suggestion that the scene sculptured

upon the

highly archaic throne of Amyklai, described by Pausanias'-^


the struggle between Menelaos

as

and Proteus,

in

reality

belonged to the wide-spread category of works of art depictThe antique


ing the combat of Herakles with a Triton.

same

sight-seers would thus have fallen into precisely the

error with respect to this representation, so entirely obsolete

French archaeologists upon

his day, as did the

in

its

redis-

covery in modern times.

Judging from the great number of archaic Greek works of


the kind which have come down to us, this victory of the
national hero over the emissary of Poseidon must have been

Gerhard

one of the most popular exploits of the cyclus.

had collected twenty-three representations of the struggle at


the time of his publication, and this list was increased to fifty
Seventeen further examples have recently

by Stephani.^

been added by Petersen,^ and,


^

finally,

seven more by Stud-

Klein (Wilhelm), Bathykles, Archdologisch-epigraphische Mittheilungen aus


In basing his argument upon
ix., Heft 2, Wien, 1S85.

OesterreichUngarn, vol.

the Olympian bronze, which forms so excellent a parallel to this representation


upon the throne of Amyklai, the author justly remarks that this change of identification removes the relief described by Pausanias from its inexplicable isolation, placing these two works side by side, at the head of, a long and typical
series of archaic designs.
2

Pausanias, III.

Gerhard, Aiiserlesene Grieckische Vasenbilder, vol.

list

is

18. 15.

not without important errors.

Thus No.

ii.

the

9,

p. 95,

note 12.

pi. xxxii.

This

of Millingen

(James V.), Peintures Antiques et InSdites de Vases Cr^cj, Rome, 1813,


showing two female figures beside the combatants.

is

incor-

rectly referred to as
*

Stephani, Compte

bourg, 1S67,
^

p. 21,

Rendu

de la Commission Imperiale Archhlogique, St. Peters-

and Nachtrog,

Petersen (Eugen),

.ffrct^/^

<?

p. 209.

Tritone, Annali,vo\.

overlooks one of the representations given

(in

Roma, 1882. Petersen


Appendix to the Compte

liv.,

the

IXVEST/CA T/O.VS A T ASSOS,


niczka,' bringing the

Among

these

we

up

total

1SS3.

to not less than seventy-four.

have, in works of

monumental

sculpture,

the poros gable, recently unearthed upon the Acropolis of

Athens,^ a bronze fragment found at Dodona,^ the Olympian bronze* and the Assos relief

Probably no single subject


mythological scenes

known

is

now under

in all the

to

consideration.

wide range of ancient

have been so frequently

illustrated.

In not one of these representations

is

Herakles shown as

attacking with his accustomed weapons, the club or the bow.


In accordance, doubtless, with
lost,

some

detail of the legend

now

the hero wrestles with the monster naked-handed, seizing

him from behind, and employing those devices of the palaisMoreover,


tra known to the sport-loving Greeks as aju/mara.
Rendu) by Stephani, to whom he refers as having collected forty-nine examples
The same remark applies to Studniczka. The example thus omitted in
only.
these recent lists is No. 416 of Dubois (L. J. J.), Catalogue des Vases Grecs formant la Collection de Mr. C. L. F. Paiickoucke, Paris, 1835.
^ Studniczka (F'ranz), Attische Porosgiebel, Mittheilujtgen des Deutsche?! Archdologischen Instituts, vol.
'-*

Described

this gable relief

the small

in the

Athen, 1S86.

xi.,

essay quoted in the foregoing note.

Other fragments of

have since been discovered, and were examined by the writer

museum upon

in

the Acropolis of Athens.

2 Carapanos (Constantin), Dodone et ses Ruines, Paris, 1S78, pi. xvi. fig. 4.
This fragmentary relief, published by Carapanos as Herakles and the Lernean
Hydra, was identified as Herakles and Triton by Studniczka in the essay before
quoted. It is a work of the second half of the fourth, or even the first half of
the third century, and is of interest as one of the very few works of so late
a date which depict this scene.
So little remains of the body of the monster
upon this fragment, that it appears scarcely sufficient to afford a conclusive
refutation of the original assumption of Carapanos
nevertheless Furtwangler
;

Roscher's Lexikon,

(in

art.

Herakles)

unquestioningly

adopts

Studniczka's

identification.
*

Aus^rabungen zu Olympia,

(Ernst),

Das

vol. iv. pi. 25, p. 19.

Engraved also

in Curtius

archaische Bronzerelief von Olympia, Abhaiuilungen der k. Aka-

demie der Wissenschaften, Berlin, 1880, p. 13, No. 6. Compare especially Furtwangler (A.), Die Bronzefunde aus Olympia und deren Kunstgeschicktliche
Bedeutung, Abhandlungen der
p. 96.

k.

Akademie der Wissenschaften, Berlin,

1880,

ARCHAEOLOGICAL INSTITUTE.

2l8

two classes of these numerous works are

by
action and grip of the

the formation of the

as well

by the

different

In the more primitive and by

hero.

Herakles

even higher.

as high as the navel, or

sits astride

monster's back, the farther leg being concealed from

view, and chokes or hugs


later

be distinguished,

as

category the fishy portion of the triton extends

far the larger

of the

to

merman

works,

him with interlocked arms.

among which

the Assos relief

is

In the

be classed,

to

these features are considerably altered, and, from an artistic

The human

point of view, improved.

portion of the fish-man

has been extended to the hips, while Herakles, shown in entire

and

figure in front of the monster, no longer bestrides

garrotes his victim, but grasps

him firmly by the

Assos, or otherwise holds him in subjection.

wrists, as at

some

It is in

measure possible to trace the development of the older works


In the most ancient representation of the

towards this type.


scene,

upon an

" island

stone "

now

Museum,^

in the British

the scales of the Triton extend quite up to the armpits, only

head and arms being human.

Herakles, on the other hand,

naked, yet with the quiver slung across his back,

is

substantially the same guise and action as at Assos.

be

difficult to find a

tency with which

more

traditions

Greeks than the identity of


;

known

the very earliest

represented by Hellenic
struggle from

we may
1

left

the

were retained by the

this type

gem

upon works

such

of

referred to being perhaps

instance of a mythological subject


art.

to right, as

Even in the movement


shown upon the Assian

of the
relief,

note the retention of an archaic feature, which

Lenormant

page 213, note

3.

(F.),

Intailles

Re-engraved

in Grieckenland, Leipzig, 1883,

in

would

It

striking illustration of the persis-

artistic

widely different epochs

shown

Archaiques de
in

V Archipel

Grec, quoted

may
above,

Milchhofer (Arthur), Die Anfdnqe der Kiinst

fig.

55.

Lenormant, as has been noted,

follows the identification of the figures as Herakles and Nereus.

representations are of extreme rarity upon

gems

of this class.

still

Mythological

INVESTIGATIOXS

AT

ASSOS,

be traced from the island stone, through

219

1SS3.

the long series

of black-figured vases, to the red-figured vase of free style

which closes the

list.^

That the advance


the changes

in

composition which

artistic

above enumerated had been made

in

led

to

Euro-

pean Greece a century or more previous to the building


of the temple of Assos,

character of

sitional

upon the Acropolis


though

still

is

the

rendered probable by the trangable

relief

discovered

recently

The hero shown upon

Athens.

of

garroting the Triton, no longer

it,

astride of

sits

the monster's back, but stands in front, entirely naked, like

There can be

the Assian Herakles.

little

doubt that these

improvements were the result of the execution


on a monumental scale.

some respects even advantageous,

in

of the vases, were

in the outline

drawings

be altogether unsuitable when en-

felt to

larged and rendered in relief^


1

of the scene

Postures which were tolerable, and

These changes, when once

Described by Brunn (H.), Viags^i in Etriiria, IV. Vasi e Speech i Chiiisini,


and engraved by Petersen in the paper before referred

BtiUettino, 1859, p. 105,


Aiinali,

to,

plays a

1882, tav.

K.

The

entire class of archaic black-figured vases dis-

marked tendency, almost amounting

to a rule, to turn

direct the action of the composition towards the right.

On

the profiles and

this point

Loschcke (G), Darstellung der Athenagebitrt, Archdologische

compare

Zeittiiig,

1S76,

Berlin, 1877.
2

Studniczka advances the contrary view.

quoted.)

It will

be remarked, however, that he

(See p. 75 of the essay before


is forced to contradict his own

theory in treating of the red-figured vase, the design of which, though depicting

an archaic subject,

is

wholly free from archaistic mannerism.

As

has been ob-

was the steady tendency


of Hellenic art to free itself from those monstrous combinations of human and
animal forms which had, in early ages, been adopted from the Orient, and to
relieve, so far as might be possible, such of these images as were retained from
Thus the extension of the human trunk
their horrid and unnatural character.
of the Triton from the armpits to the waist, as in the Attic gable, the Assos
The assumption that
relief, and the red-figured vase, is a distinct advance.
these figures appertain to a more ancient type than that of the black-figured
served

in

connection with the human-legged centaurs,

vases, or of the island stone,


criticism.

is

it

at variance with leading principles of historic

ARCH^OLOGICAL INSTITUTE.

20

made, were readily adopted by designers of all classes, as is


shown by the painting of the red-figured vase before referred
In view of the fragmentary condition of the Athenian
to.
gable,

it

is

how

determine

difficult to

far

the Assian relief

That the provincial

was

directly influenced thereby.

who

decorated our temple depended largely upon prominent

works

of

Attica and the Peloponnesos

ia

certainly to

assumed, as we have had occasion to observe


difference the

interlocked arms

adoption

of the

may possibly,

in

wrist grip

be

connection

The main

with the relief of Herakles and Pholos,

artists

point of

instead of the

as Petersen has suggested,

be ascribed to a misunderstanding of the wrestling hold of


the older type, through which the arms of the combatants

were interchanged.

The

readiness with which

this

might

happen may be judged from a comparison of the illustration of the relief. Fig. 49, with the drawing of a blackfigured vase. Fig. 50.

Yet the present writer

adopt this explanation,

in

view

of

hesitates to

the otherwise intelligent

character of the design, and the evident purpose of the sculp1

The observations of Petersen (Essay

in the

Annali, 1S82, before referred to)

in regard to this relief are so interesting as to deserve quotation in full:


d'

uopo gettare uno sguardo

di confronto

anche sul

rilievo

d'Assos.

Fu

"
gia

osservato che la forma di Tritone e quivi analoga a quella della tazza a figure

R' [published as plate K of the Amiali, 1882]. Con lo stesso \si'c\ e con
esso ha comuni le donne che corron via. II gruppo dei combattenti
poi non mostra mai altrove deviazioni si notevoli dal tipo antico se non in
'R'; ma vi contribul forse anche la necessita dello spazio. Ercole non cavalca pill Tritone vedesi pero ancora serbato il gran passo; egli non abbraccia
piu Tritone, ma gli afferra le braccia, non si sa bene con quale scopo.
Chi
sa che 1' autore di esso rilievo non abbia frainteso il tipo Greco, come pure e
rosse

'

pochi

altri

accaduto a recenti osservatori, ed abbia scambiato il braccio destro di Tritone


con quello di Ercole ? Ammesso questo, e imaginandoci che la sinistra di Ercole
abbracciasse

anche

avrebbe

il

petto di Tritone e fosse quindi intrecciata

la sinistra alzata di

affinita

rilievo."

il

suo esemplare nei

con

'

R,'

ma

coll'

altra,

allora

Tritone e stringente senz' alcun dubbio un pesce


vasi.

L'

ammetter questo abbaglio diminuirebbe

giovcrebbe a mettere in luce

il

carattere semigrcco del

INVESTIGATIONS AT ASSOS,
bestow upon the Triton his

tor to

attribute,

22

1SS3.

the

conch held

hand.

in the uplifted right

For despite the weathered and battered surface of the


and the consequent uncertainty of outline, there can

rehcf,

be

doubt that

little

this object

through which the distressed

No

for aid.
rac,

De

actually a shell trumpet,

merman

about to blow a

is

call

explanation was offered by Texier, Cla-

Witte, and Guigniault, the

sculptures,

mind

dififerent

is

and no other

four editors of these

first

possibility suggested

itself

the

to

of the present writer, during a close scrutiny of this

relief.

Yet Stark ^ speaks

sees in

it

of the object as a ring;

from Nereus

and Petersen and Wolters

have recently pub-

as their opinion, that the attribute

lished

it

but a

fish,

Tiimpel^

a handle of the sun-bowl which Herakles received

seemingly basing

quent appearance of a

fish in

this

not a conch,
fre-

other representations of Tritons

than upon an examination of the work


Studniczka,* following the

is

view rather upon the

same method

Furthermore,

itself.

of determination,

and

arguing from an ancient description of a certain " statue


a Triton, carved of
its

wood and holding

De

Villefosse,

present Director of the Louvre, in reply to inquiries


Purgold, that the attribute in question

He

hand,"^ believes the object to be a drinking-vessel.

quotes in this regard the statement of

fish,

is

zmd die

the

made by

at all events not a

and may be either a drinking-horn or a conch.

Stark (C. B.), Gaza

of

a silver kratanion in

Were

Philistdische Kiiste, Jena, 1852.

Tiimpel, Die Aithiopenldnder des Androtnedamythiis, Supplementband der


yahrbiicher fiir dassische Philolo^ie, Leipzig, 1887, p. 199.
'

Friedrichs and Wolters, Die Gypsabgiisse antiker Bildwerke in Jtistorischer

Fol^e erkldrt, Berlin, 1885.


*

Essay

in the

Mittheilimge7i des Detitschen Archdologischen Instilutcs,

before quoted, p. 67, note

1S86,

i.

6 Polemon (or the author of the Matiners and Customs


of the Greeks) quoted
by Athenaios, XI. 59, p. 480a. The Triton thus described, evidently a xoanon,

stood in the treasure house of the Byzantines at Olympia.

ARCH^OLOGICAL INSTITUTE.

22 2

necessary to bring forward literary and archaeological

it

guments

to explain

of a Triton,

it

ar-

the appearance of a shell in the hand

would not be

on the one hand,

difficult,

adduce a great number of classic references to

to

this attribute,^

while, on the other, besides the ancient account of figures of

Tritons blowing conches in the gable of a temple far older

than that of Assos,^ we have examples of the kind upon well

known

coins of Corinth, Akragas, and Askalon, as well as in

vase paintings, and on


too

numerous

reliefs,

to mention.^

engraved gems, mosaics,

It is

worthy

great majority of these representations

of

show the conch

a curved form, like that of the Assian

etc.,

remark, that the


as of

relief.

Recent investigations have given us much information concerning the origin of this monster, graphically designated

by an inscription upon the archaic Olympian bronze as the

"Old Man

Herakles, he

of Oriental, and, as

is

The

extraction.*

it

Like his adversary,


appears, Phoenician

similarity of the formation of this fish-man

of the deities

that

to

Sea" (AAI02 TEPfiN).

of the

Dagon and Derketo,

especially wor-

shipped in Askalon and Gaza,^ must strike every observer.

The prototypes
1

of the

Pausanias (VIII.

2.

Greek Triton appear upon Assyrian

7) states as particularly characteristic of Tritons that

Compare

they blow through perforated shells.

Nonnos, Dionys.,

I.

61, VI. 270,

also Moschos, Idyll., II. 120;


93; Virgil, Aen., X. 209; Lucan,
Pliny, Hist. Nat, IX. 9; Hyginus, Fab.,

XXXVI.

Mdam., I. 333
Appuleius, Alftam., IV. 85; and many others.
This was the temple of Saturn, in Rome, destroyed by

Phars., IX. 347; Ovid,

II. 23;
2

the city 257.


^

list,

conches,

is

The account
far

referred to

is

given by Macrobius,

fire in

from complete, of such representations of Tritons and their

given by Stephani in the Compte Rendu de la Co7nmissio7i Imperiale

Archeologiqiie, St. Petersbourg, 187

1.

Compare, upon the course of this development, the remarks


Anfdnge der G'^iechischen A'unst, p. 84.
^

the year of

Sat., I. 8. 4.

discussion of this point

examples, see Lajard

(J.

is

to

be found in Stark, Gaza,

of Milchhofer,

p. 249.

For

B. F.), Recherches sur le Culte, les Symboles, Ics Attri-

buts et les Mouiimettts figurh de Venus, en Orietit et en Occident, Paris, 1837-49,


pi. 22, 24.

INVESTIGATIONS
and seals

reliefs

and

;^

termed by keramic

AT

ASSOS,

upon that

of painted vases

class

specialists of the older school Tyrenian,

we

Phoenician-Babylonian, or Syrian-Phoenician,^
the offspring

of similar shape,

sters

tradition.

223

1SS3.

Furtwangler^ believes

image

this

see

same

the

of

to

mon-

artistic

have been

introduced to Hellenic art by the Ionian inhabitants of the


Asiatic coast during the eighth century before Christ.
tain

it

is

Cer-

that the island stone, before referred to as repre-

senting the struggle of Herakles with the monster, shows


distinct traces of the Phoenician influence.

The

Attic vase painters

tions of this

who

inscribed their representa" Triton "

monster with the name

undoubtedly

made by Greek
The image of the

followed an identification of this Oriental type

mythographers

of a

much

earlier age.

Phoenician sea-god was readily

member

corresponding
esting

even

to

observe, as

made

of the Hellenic theogony.


illustrative

at the early date

when

of the

It is

inter-

Greek conception,

this adoption

such a monstrous combination of


1

to serve for that of

took place, that

human and animal forms

As, for instance, the relief from the palace of Sargon, which dates from the
and Flandin, Momaiieiit de Ninive,

close of the eighth century, engraved in Botta


Paris, 1S49-50, pi. 32, 34.
in

Lajard, Mithra,

Babylonian seals representing the subject are shown


17, 2; 31, 5; Layard, Discoveries in the Ruins of

pi. 62, i, 2;

Nineveh a?id Babylon, London, 1853, p. 343; King (Charles William), Antique
Gems a7id Rings, London, 1872, vol. ii. pi. 3, 6; and others.
2 On this class of vases compare De Witte, Cabinet Durand, Pref. IL and
IIL; Gerhard, Archdologisches Intelligenzblatt, Berlin, 1836, p. 307; Raoul Ro(^^\Xt, Noiivelles Observations sur les Anciennes Fabriqiics de Vases Feints,

Journal

des Savants, Paris, 184 1, p. 356.

Human figures

ending

in fishes' tails are

shown upon

the so called Phoenician

vase published by Gerhard, Ueber die Kiinst der Phdnicier, Berlin, 184S, pi. 47
and upon those given by the same author in his Berlm's Antike Bildwerke,
;

Nos. 480 and 542.


Micali,

Momimenti

Upon

a vase

now

in the collection of

Inediti, Firenze, 1844, pi. 43,

Munich, published by

sphinxes and

griffins closely re-

sembling those of Assos appear, together with a winged figure provided with a
similar appendage.
^

Furtwangler, Die Bronzefu7tde aus Olympia, pp. 96, 97.

ARCH^OLOGICAL INSTITUTE.

2 24

was never employed


self,

beings, the

of

The name

as

forms as Amphitrite

restricted to an inferior order

first,

emissaries of the

frightful

Triton,

mighty Poseidon him-

to represent the

but was, from the

we may observe

Earth-Encircler.

such compound

in

and Tritogeneia,^

directly signifi-

is

cant of the element inhabited by these creatures, having, as

has been recently pointed

out, a

common

origin with the

Sanscrit word trita, water.

Pausanias

has given us a description of the appearance of

these beings, which corresponds closely with their portraits

upon the archaic works

of

art,

and furthermore indicates

various details which upon the reliefs of

were undoubtedly supplied by


hair

color.

resembled the frog-grass seen

Athens and Assos

He
in

remarks that their

swamps,

falling

in

masses so that the separate hairs were not distinguishable;


that they had a

human

fingers indicated

nose, eyes of a bluish tint, hands with

and with finger-nails similar

and that they had below the


a

tail

and

feet,

like that of a dolphin.

When

the conception

Triton,^ as
^

to mussel shells,

belly, instead of legs

is

well

It is interesting to

is

confined to a single individual, this

known, takes
observe

how

and the true


the Lake Tritonis

his place in the

entirely the

mythology

Greeks themselves were

rant of the derivation

significance of this word, referring

two syllables

in Libya, near

its

of

ignofirst

which Athena was born,


to the stream Triton in Boiotia,
or to the numeral Tpets,
to the head of Zeus,
either because the goddess was born on the third day of the month, or was the
third child born (after Apollo and Artemis), or, finally, was the author of the
three main bonds of social life. The references to the classic authors who have
to

thus explained the word will be found

Compare upon

in

Stephani's,

Thesauros,

s.

v.

Tpira-

Welcker, Die Aeschylische Trilogie, Darmstadt, 1824, pp. 164, 2S2.


The recognition of the Sanscrit root can leave no doubt
as to the real significance of the epithet, of which not one of the ancients seems
to have been aware.
yiveia.

Pausanias, IX. 21.

Triton

is

this point also

i.

said to have been the son of Poseidon by Amphitrite (Hesiod,

Thepg,()2P, and Apollodoros,

Lykophron,

8S5), or

I. 4. 6),

or by Kelaino (Tzetzes, commentary to

by Salakia (Servius, commentary to

Virgil, Aen.,

I.

144).

AT

INVESTIGATIONS

ASSOS,

18S3.

225

the Greeks as a son of Poseidon, dwelling with his parent in

a golden palace beneath the waves.^

But why

this son or satellite of the sea

is

god so frequently

depicted as thus struggling in the rough embrace of Herakles

Not one of the archaeologists who have

treated of the

subject has had the slightest explanation to advance in this regard.

VVelcker, Gerhard, and Stephani alike remark that the

make no mention whatever

ancient authors

popular as

was

in the earliest ages of

of such a combat,

Greek

art.
Baumeisand Furtwangler speak of the legend as altogether unattested by the mythographers.
In short, all those who have
it

tcr

treated of the

subject are

ma

tazioni,

in

agreement with the concise

" abbiamo frequentissime rappresen-

conclusion of Petersen,

nessiina viensione nei siiperstiti momimeiiti

let-

terariiy

Yet

spite of this great weight of authority, the present

in

writer believes

and

gle,

it

possible to explain the nature of the strug-

to connect this large class of ancient

works of

art

with one of the most notable exploits of Herakles, recounted

by

is,

known

that

sea monster
the

life

to

who devastated

Thus described

the foregoing note.

was situated

which he

these very coasts and threatened

of Hesione, in visitation of the

This legend

at

combat represented is
have taken place between Herakles and the

briefly stated, that the

Laomedon, king

The conclusion

classic authors of every age.

has arrived

wrath of Poseidon upon

of Troy/'^
is

one

in the

of

the oldest of the Trojan

Cyclus,

passages of Hesiod and Apollodoros referred to in


to the Homeric idea {Iliad, XIII. 20), this palace

According

at Aigas,

the

name

derived, without doubt, from the

of half a dozen

same root

Greek towns near the


and oX-yiaXis.

sea,

as At^o'cof

- This explanation was suggested in the Prelimifiary Report,


the
p. 106,
ground for the reason there assigned being the local character of the Hesione

legend.

advanced

It is

perhaps

at that

fair to state that the identification would not have been


time had the writer then been aware of the difficulty of sup-

porting this bold hypothesis in the lack of


15

many arguments

since collected.

ARCHAEOLOGICAL LXSTITC/TE.

2 26

connected with the most primitive traditional history of the


country,

and repeatedly referred

Homeric

epics as

familiar to

if

by the singer

to
all

his

hearers.

the

of

Poseidon,

together with Apollo, had been bound over to serve Laomedon, king of Ilion, for a
a revolt against the
to test the

full

power

year,

whether

presumptuousness of

this mortal ruler.^

this year, Poseidon, obviously in

Laomedon

During

his character as Asphalios,

When

built the fortification walls of the city.

performed,

punishment for

in

of Zeus, or voluntarily, in order

was

this task

refused to give the gods the wages

which had been promised them, and drove them from his
dominions, threatening to cut

hand and

foot,

sell

them

to bind

their ears,
in

some

them

distant island as

In revenge for this ignominious treatment Poseidon

slaves.

a sea

sent

and to

off

who ventured

monster, which destroyed those

upon the sea-shore, and even those whom it caught tilling


Laomedon, in distress at the suf-

the fields near the coast. ^

had thus befallen

fering which

oracle of Apollo for a remedy,

must be

sacrificed to the

his

people, inquired of the

and was

that a virgin

told

monster as a propitiatory

offering.

upon Hesione, daughter of the king, who was


accordingly exposed to her fate upon the promontory of
Agamias or Agammeia,^
a spot to-day recognizable in the
steep and desolate point of land which forms the northern

The

lot fell

Iliad,

VII. 452 and

XXI.

443.

Apollodoros,

II. 5.

9; Hesiod, quoted by

Horace, Car7n., III. 3. 21 Valerius Flaccus,


the scholiast to Lykophron, 393
Argon., II. 491 ; Servius, commentary to Virgil, Aen., II. 610.
;

Of

this

vengeance, the fullest account, and that preserving most of the

tures of the archaic legend,

is

For other

given by Diodoros, IV. 42, 49

fea-

refer-

ences to the story of Herakles and Hesione see Lykophron, Cass., 34, with the
commentary of Tzetzes; the scholiast to the Iliad, XX. 145; Apollodoros, II

Dictys of Crete, IV. 22 Philostratos Jr., Imag., 13 Eudokia, Viol., p. 344;


5. 9
Isaac Porphyrogenitos, preserved in Allacci (Leone), Excerpta Varia, Romae,
1641, p. 272; Hyginus, Fab., 89; Valerius Flaccus, Argon., II. 497-533; Servius,

commentary
8

to Virgil, Aen.,

Hesychios,

s.

I.

v. 'Ayajxlas,

550, III. 3, VIII. 157.

and Stephen of Byzantion,

s.

v,

'

Ay dfd.fj.e to.

lAVESTICATIONS

AT

ASSOS,

22/

1SS3.

Beshika Bay. At this moment it chanced that


Herakles, proceeding on his expedition against the Amazons,
passed that way, and, releasing Hesione, took her with him

boundary

of

Here Laomedon induced the hero

into the city.

to

go forth

to battle against the monster, offering as a reward, in case


of success, the immortal horses which his grandfather, Tros,

had received from Zeus in compensation for Ganymede. The


The only
details of the struggle have been variously related.
archaic account, that of

Homer,

tells of

a wall of earth which

the Trojans, with the help of Athena, piled up for the protection of the hero

"when he

the coast to the fields."

should be driven back from

Other and

later descriptions

the encounter will subsequently be referred

of

to.

Setting aside that late imitation of the Homeric story of

Menelaos and Proteus, by which Herakles was connected


with Nereus,

a legend excluded, moreover, from the present

consideration by inscriptions upon archaic vases, as before

mentioned,
ster sent

this exploit of the

hero

in

by Poseidon against Laomedon

subduing the monis

the only feat of

the kind which the ancients attributed to Herakles.

be no other than

this

deed

to

which Euripides

refers,

It

can

when,

in

connection with the twelve labors, he describes Herakles as

niad,

XX.

The same statement in regard to the wall is made by the


who refers to Hellanikos as having related the story.
Jason der Drachentodter, Rkeiiiisches Museum fiir Philologie, etc.,

145.

scholiast to this passage,

Welcker

(F. G.),

Bonn, 1835,

III.,

tingen, 1S49-64,

subsequently reprinted in his Alte Denkmdler erJddrt, Gotby Wieseler (Friedrich), Herakles in den Rachen des

followed

Meeriingeheuers tretend
schaft, Giessen, 1851,

tenbilder,

MUnchen,

und

die hefreite Hesione, Zeitschrift fiir Altertlnmiswissen-

Nos. 40 and

1S70.

41,

and Flasch (Adam), Angcbliche Argonau-

emends the scholion by substituting t^vxos for

ret'xos,

was supplied by his patroness with means


But in view of the clear account
of attack rather than with means of defence.
that the
of the wall and its purpose given in the Homeric text, and of the fact
word Tei'xos is repeated by Tzetzes (commentary to Lykophron, 34) this change
preferring to believe that Herakles

appears altogether inadmissible.

ARCHAEOLOGICAL INSTITUTE.

28

bay of the sea and establishing with his oars a


or which is meant by Pindar,^ Sophomortals,"^

" entering a

calm

for

kles,^

and Euripides

another passage,* when they speak

in

of the hero as clearing the sea of

its

monsters.

would, indeed, be an altogether unparalleled case

It

mythology

of art,

known

in

the

a deed so celebrated as to have formed

some seventy

the subject of
ings

if

to us should

of the black-figured vase paint-

have been passed by entirely without

mention by the ancient poets, playwrights, and mythograBut here we have full accounts of a legend, of exphers.
the highest antiquity, which

ceptional popularity and of

applicable to this scene, or

by Greek

The

is in its

art.

fact that this identification has not hitherto

posed

is

been pro-

undoubtedly due to two considerations, which

suggest themselves to every archaeologist as objections.


these

first of

was

called

cially in

is

turn wholly unrepresented

is

that the marine monster sent by Poseidon

by a word

its

will

The

(/ct^'to?)

which

in later ages,

and espe-

Latin form, gradually came to be restricted to

large sea animals having an actual existence, such as whales,


sharks,
this

among
1

and the

change of
the

like

idea,

that,

in

consequence

of ancient art of a later period

Romans, though not among the Greeks

Euripides, Here. Fur., 399

TiBeU

the second,

works

irocrias

0'

of

actually

aAhs ixvxoi>s elcrfPaiye, Qvaro'is yaXavelas

ipiTfxois.

This passage shows that the exploit was not performed with the intention of
forcing the sea-monster to prophesy, but rather for the purpose of establishing
peace for mortals. The figurative words of the poet fully characterize the deed
as the deliverance of
^

some human
;

sufferer.

and Nem.,
Pindar, Isthm., III. 75
Sophokles, Track., 1012.
Euripides, Here. Fur., 225.

I.

62.

These references are

in entire contradiction

with the only explanation of the purpose of the combat between Herakles and
Triton hitherto advanced, namely, that of Furtwangler (in Roscher's Lexikon,
art. Herakles, section iii. p. 2192), who speaks of the Triton as "subdued and
held in restraint until he imparts his secret knowledge to Herakles."

AT

INVESTIGATIONS

ASSOS,

229

1SS3,

represented the monster to which Hesione was exposed as


a large fish or sea- dragon.
That, however, the word Ketos, as employed by Homer,

not necessarily to be taken to

mean

a fish of any kind,

dent from the context of other passages,

is

is evi-

which the word

in

applied to sea-dwelling monsters of entirely different na-

is

ture.^

Elsewhere, the word

as crocodiles

is

used

such frightful beasts

for

And,

and hippopotami.^

finally, that

himself considered to be one of these monsters

is

Triton was

evident from

the reference of Lykophron to him as Poseidon's


Kvdiv? to which the scholiast adds, that this Triton
tos

subdued by Herakles.*

This passage

may be added another

To

and most curious instance

employment

am

it

which attention has

this signification, to

aware, hitherto been drawn.

It

the Ke-

conclusive, and

is

completes the chain of evidence.


of the

Kdp')(^apo^
is

of the

word

not, in so far as

in
I

appears, namely, from

the accounts of Demostratos^ and Pausanias,^ that a show

monster, purporting to be the embalmed body of a Triton,

formed one

of the chief sights of a

same way

precisely the

as the

temple

mummies

at

Tanagra,

in

mermaidens and

of

sea-serpents are exhibited in the booths of country fairs at


the present
1

The word

So famous was

day.

this

ancient curiosity,

for instance, applied to the sea-calves

(seals) of Proteus
"sea-beasts" by Butcher
their version of the Odyssey, and by Voss, still more correctly,
is,

{Odyssey, IV. 446, 552).

It

is

correctly rendered

and Lang in
"Meerscheusal"; but Buckley (London, 1S80) translates the word "whales,"
absurdly mistaking the obvious sense of the passage.

The

translator last

named

thus exemplifies the error into which the mythographers of later antiquity had

themselves been

led.

Euthymenes, quoted by Athenaios, II. 90.


Lykophron, Cass., 34, with the scholion of Tzetzes to

Preller

[Griechische Mythologie, vol.

passage of Lykophron quoted


longs to the category of the
6

Demostratos,

Pausanias, IX. 20.

in the text,

p. 163,

note

this passage.
2),

in referring to the

remarks concisely, "Triton also be-

jc^ttj."

Halieictics,
4.

ii.

preserved in Aelian,

De

Animal., XIII.

21.

ARCH^OLOGICAL INSTITUTE.

230
that its

was frequently impressed upon the coins

effigy

of

Tanagra as one of the symbols characteristic

of the

The bloated appearance of the stuffed animal


came proverbial; and Athenaios, in adducing it

evidently be-

town.^

as an illustra-

tion of excessive obesity, has referred to the defunct Triton

as "the

Ketos

of

Tanagra,"

thus leaving no doubt as to the

direct applicability of this term to the being in question.

The instance
of Hesione

is

by the adventure

of his evil deeds afforded

what the ancients have

quite in accord with

otherwise reported in regard to the nature of Triton.

Tanagra he was believed to have attacked the


went down

to the sea to bathe,

and

to

At

women who

have carried

off

the

herds grazing near the coast,^ even as he had at Troy in

King Laomedon. And without doubt such a


Triton was in the mind of Odysseus, when he dreaded lest
the days of

Poseidon should send an


It

is

evil

Ketos against him.*

nevertheless undeniable, that, through

restriction of the

word

to

members

the

gradual

of the fishy tribe, the

conception of the monster to which Hesione was exposed

came more and more to resemble a whale, rather than a merman. Thus the scholiast to the Venetian manuscript of the
IHad ^ relates that Herakles entered the body of the sea beast
1

number

of such coins have

been collected by Wolters, Dei- Triton von

Tanagra, Archdologische Zeitung, 18S5.


^

Athenaios, XII. 75,

irevSiJ.evov

Tovvras

Tw

elvai

p.

Aeirr^Tepof

551 A:

Xlidtf ohv

uy KaraXeyet

Tava"ypau) kt|Ti ioiKivat, Kaddirep

in his edition of the

Deipnosophists

Ka\\i6u

"EpfjLiinros
01

icrriv,

iu

irpoeipTf/j.^vot

(vol. iv. p. 253),

ayaOe TijxdKpaTes,
^

vwfpTrXov-

&vSpes.

Meineke,

KepKcoxpiv

has questioned the correct-

ness of this passage, basing his doubts upon a corrupt gloss of Hesychios.

question raised has been adequately discussed by Wolters

in

The

the essay quoted

in the foregoing note.


s

Pausanias, IX. 20.

4.

This legend appears, indeed, to be a duplication of

the Hesione story.


*

Odyssey, V. 421.

XX. 146. The commentator remarks that the story of


Herakles and Hesione had been related by Hellanikos, but it by no means fol^

.Scholiast to //iad,

INVESTIGATIONS
and pierced

its

ribs; while

AT

ASSOS,

23 I

1SS3.

Theophylaktos^ and Tzetzes,^ writ-

ing in the eleventh and twelfth Christian centuries, go

still

farther in this direction, asserting, in obvious imitation of the


Biblical legend of Jonah, that the hero

monster

of the

remained

in the belly

That the archaic

for the space of three days.

Herakles having wrestled with the monster

conception of

bare handed had been entirely forgotten, even in the classic


period of

Latin literature,

may

be gathered from the long

account of the combat given by Valerius Flaccus,^ who describes

the hero as killing the monster with a rock, after

vainly attempting to

wound

it

And

with arrows.

that

some

uncertainty in this respect began to creep into the legend

may be

early in the epoch of the red-figured vases of Greece

surmised from the

fact, that,

upon the single representation

of the subject belonging to this category, Herakles

is

shown

armed, although not attacking, with his club.

Probably no other ancient legend became, as time went


on,

more

entirely perverted.

The

later writers

confounded

the deliverance of Hesione with that of Andromeda, the Triton


with the dragon and with Jonah's whale, and the tactics of

The ven-

Herakles with those of Menestratos the Thespian.^

geance wrought by mighty Poseidon upon the presumptuous

Laomedon and

his subjects appears, in the account preserved

by Diodoros, as the

God who destroyed

visitation of a divine emissary,

of the

Sea

the people by a plague, and blasted

the fruits of the field."

what

"a

all

This original conception had some-

mystical character of

from which

the Oriental mythology

it was derived
its vagueness rendered it the
more superhuman and terrible.
But when this Sea God
;

lows from this fact that

all

the details of the tale are to be referred to the

more

ancient authority.
^

Theophy\aktos, Ex/>ositio in ProJ>/ie(am Jonam, cap.


Scholion to Lykophron, 34.
Valerius Flaccus,

II.

497-533-

ii.

(ed.

Migne,

Pausanias, IX. 26.

5.

p. 1S9).

archjEological institute.

232

Greek

reappears, in the decrepitude of

idea

is

wholly

lost,

culture, the archaic

and we see nothing but an ugly beast

the hungry, carnal monster depicted in the shape of a conventional dragon.

Thus

it

came about

Roman

that in

when

art,

the story of

the exploit was again popularized and freely paraphrased by

Latin mythographers, the Ketos became at


cean, or sea-dragon.

vase and the

Dodona

relief,

any work of ancient

in

When,

tian era.^

we do

This

is

a true ceta-

not find this scene depicted

art until after the

advent of the Chris-

after this long break, the subject

taken up, the change has been


1

last

Subsequent to the isolated red-figured

No

fully effected.

less

is

again

than six

a highly remarkable fact, for the story of Herakles and Hesione

was, in the third century before Christ,

still

sufficiently

popular to have been

chosen as the subject of a comedy, called Hesione, by the poet Alexis. Ribbeck (Otto), Die Rdmische Tragodie ini Zeitalter der Republik, Leipzig, 1S75,
p. 46, is thus in error when he asserts that no Greek drama is known to have
From the fragment of the play preserved by Athenaios
treated of this legend.
(XI. 41), we may see that the struggle with the monster was related in detail,
the lines in question describing the exhaustion and great thirst of the hero after
the exploit.

From two of Pliny's lists {Nat. Hist., XXXV. 114 and 139) we know that
the subject was treated by Antiphilos and Artemon, artists of the Hellenistic
epoch, but the mention throws no light whatever upon the nature of these repreThat the exploit was not forgotten in the subsequent ages would
upon an Apulian amphora in the Museum
of Berlin, if we are to accept the explanation given by Gerhard {Apulische Vasenbilher, Berlin, 1845, pi. xi., described in the same author's Berliii's Aittike
sentations.

likewise be proved by the painting

Bildwerke, No. 1018),

who

identifies the figures as

Herakles, after the struggle

the rescued, yet still


I.aomedon to ask for Ws reward,
This vase has however been referred
fettered Hesione following with Telamon.
by Furtw'angler (Beschreilnnig der Vaseiisamnilim:^, No. 3240) to an entirely different subject, Laomedon appearing in this identification as Kreon, and Hesione
Compare the critical literature to the subject cited in the volume
as Antigone.
is

over, approaching

last

mentioned.

likewise uncertain whether we should include among these examples the


painting upon an Etruscan vase, found at Perugia, of late yet fine style, engraved in the RIonumcnti, vol. v., Roma, 1849-53, pi. ix., in the Anttali, vol. xxi.
pi. A, and in Welcker's Alte Denkmdler, vol. iii. pi. 24 (compare the reIt is

1849,

marks

in the

Adunanze,

Bullettino, 1846, p. 87),

which depicts a hero advancing

AT

INVESTIGATIONS
of the

Pompeian

ASSOS,

1SS3.

wall-paintings, hitherto unearthed, represent

the deliverance of Hesione, and in those in which


distinguishable the monster

without

human

233

is

semblance.-^

its

form

is

a veritable sea-serpent, wholly

In two of these instances

it

is

mouth of an enormous fish or dragon. This scene


was identified by Wclcker {Wilson dcr Draclientddtcr, quoted above, p. 227, note),
and by Emil Braun {Ingresso di Giasotte iidle Fatici del Dragone, Aiinali, vol.
xxi., 1S49), as Jason and the dragon, but has since been held by Wieseler (Iferakles in dot Rachen dcs Rlccrungcheiters trdoid, before quoted) and by Flasch
(Aiigebliche Argonaiitcn Bildcr, III.) tO represent Herakles and the sea monster.
Flasch even goes so far as to detect in the mantle which the hero has thrown
over his head the rei/xos of Welcker's emendation, before referred to. Although
there is nothing whatever upon the vase peculiarly characteristic of Herawith drawn sword into the

kles, recent writers, e.g. Baumeister, Dcitkmdler, Art. Herakles, (Furtwangler,


in

Reseller's Lexi/coti,

identification, which,

Art. Herakles,

were

vase-painting the earliest

it

less

is

committal,) incline to the latter

susceptible of proof,

known

would render

this

Etruscan

representation of that version of the legend

which is given by the scholiast to the Iliad.


1 No. I.
Reale Accademia Ercolanese di Areheologia, J^itlure Antiche d'ErcoHelbig (Wolfgang), Wandgemdlde der vom
lano, Napoli, 1757-92, vol. iv. p. 62
A hero upon
Vestiv verscJiiitteten stddle Canipaniens, Leipzig, 1S6S, No. 1129.
the shore, identified as Herakles by Wieseler, but as Telamon by the Neapolitan
Academicians and by Helbig, hurls a rock at the monster. In the background
another hero armed with a club (Telamon according to the former view, Herakles according to the latter) converses with two women, one of whom, naked,
Notwithstanding the indorsement of so high an authority
is doubtless Hesione.
as Helbig, the view of the Academicians, which would attribute the destruction
;

of the monster to Telamon, appears altogether inadmissible.

The monster is slain by a rock in


This fresco had been identified by the Academicians as Perseus
and Andromeda, but was correctly explained by Wieseler, to whom Helbig has
No.

like

2.

Helbig, Wandgemdlde, No. 1130.

manner.

adhered.

No.

3.

Helbig, Wandgemdlde, No. 1131.

Herakles armed with the club;

the image of the monster defaced.

No.

4.

Helbig, Wandgem'dlde, No. 1132,

Compare Schone, Scavi

pi. xiv.

Herakles standing as victor, armed with club


and bow, while Telamon releases Hesione with a hammer from the shackles
which bind her to the rocks.
No. 5. Helbig, ^Fa^i^^na/i/^, Appendix, p. 458; Sogliano (Antonio), /'///wr^'
Murali Campane, Napoli, 1S80, No. 494 Kekule, Scavi di Pompei, Bullettino,
Herakles armed with bow and club the monster entirely, and
1S67, p. 165.

di Pompei, Bullettino, 1867, p. S3.

Hesione

No.

6.

in greater part defaced.

Robert, Adunanze delV Instituto, Bullettino, 1S75,

p. 40, identifies as

ARCH^OLOGICAL INSTITUTE.

34

killed
third,

by a rock, as described by Valerius Flaccus ^ in the


while in the others Herakles
it lies pierced by an arrow
;

armed with club and bow. We may perceive from this


variety, among works nearly contemporaneous, how freely the
details of the legend had been treated by mythographers of
is

how

the period, and

come

entirely the original conception

had be-

In conformity with the account of the exploit

obsolete.

who had at this very time brought the


vogue, Telamon appears as the companion

given by Hyginus,^
story again into

of Herakles in two of these frescos, as he does


later

terra-cotta relief of the scene engraved

and upon the well known mosaic

upon the

by Campana^

of the Villa Albani, published

by Winckelmann.*

Upon

Roman workmanship,

belonging to Gerhard and published by

him,^ Herakles bends his

a fragmentary

bow

paste

cameo, of

against the dragon guarding

Hesione, the monster having here the head of a serpent


rather than that of a

marble discus

in

fish.

the

A relief

Museum

of

of

debased style upon a

Vienne

likewise

shows

the struggle to have been carried on with the bow.


Herakles and Hesione the painting given in the Pitttire d'Ercolano, vol. iv. p. 6t,
and in Helbig, Wandge77idlde, No. 1184, where it had in both cases been described as Perseus and Andromeda. The hero wades into the water to meet
the monster.

As proof of the popularity of the subject


may be added to these pictures of the

there

early in the

first

Christian century,

deliverance of Hesione the presen-

Priam to Herakles by Hesione (Helbig, Wandgemdlde, No. 1147, upon


which identification compare the extensive literature there quoted).
tation of

1
'^

Valerius Flaccus, Argon., H. 533


Hyginus, Fab., 89.

Campana

" Alcides

saxo surgentia colla obruit."

(Giovanni Pietro), Atttiche Operein Plastica, Roma, 1842-52,

pi. xxi.

Winckelmann, Monumenti Antichi, Roma', 1767, vol. i. pi. 66. Engraved


also by Guignault, Religions de V Antiqtiiti, pi. clxxxii. No. 663
and by Millin,
Galerie Mythologique, pi. cxv. No. 443.
5 Gerhard, Gemmenbilder, Archdolooische Zeitting, 1849,
This is,
P^- '^'' ^o- 4*

presume, identical with the

gem

representing the subject, referred to as

edited in Gerhard's Afulisc/ie Vnse7ibilder, p. iS.


<

Stark (K. B.), Mtiseographisches, Archdologischer Attzeiger, 1S53, No.

52.

in-

INVESTIGATIOXS

AT

ASSOS,

235

18S3.

appears to have been an immediate result of the rep-

It

story by Latin

of the

etition

that

poets,

this

dehvery of

the Trojan princess was adopted by the decorative

the

Romans

most characteristic

as one of the

of

art

exploits of the

heroic age, and perhaps even regarded with a certain naorigin of the earliest rulers of

tional pride in the mythical

It was introduced, obviously in this significance,


Latium.
and as typical of the valor of Trajan, that second Herakles,
among the reliefs of the Triumphal Arch at Treves, fragments of which have recently been brought to light.^ That

the legend retained

its

latest ages of antiquity

Roman

ance upon a
to the

place in popular favor until the very

furthermore proved by

is

sarcophagus, referred by

beginning of the third century of our

Museum

served in the

Cologne

of

i^

era,

by an

its

appear-

inscriptions

which

is

pre-

altar of similar

character in the palace garden of Durlach;^ and by a fragment of a rude sandstone relief also discovered in the Rhenish
Provinces,^ which shows the figure of the chained Hesione.
N. von), Die rbmische Villa zn Nennig, fig 4. Trier, 1868.
Sarkophag im Musaun zti Kobi, Jahrbuch des Vereins von
reAlterthiunsfreiinden im Rheinhmde, vol. vii., Bonn, 1845, Plates iii. and iv.
published in the same author's Alie Denkmaler, vol ii. Wieseler, in the paper
1

Wilmowsky

Welcker

(J.

(F. G.),

far Alterthiimswissenschaft, 1 851, before quoted, has pointed out


hand of Herakles is the stone which plays a part
not an apple, as supposed by
in the story as told by Valerius Flaccus,
Welcker ni his contemptuous remarks concerning the sculptor.
8 Urlichs, Neiister Zuwucks des k. Museums, Jahrbuch des Vereins von Alterin the Zeitschrift

that the object held in the

thumsfreiinden im Rheinla7tde, vol.


*

Engraved

Upon
kles

is

in the

Roman

shown

sarcophagus

forcing

ix.,

paper referred to

the Villa Borghese, apparently inedited, Hera-

open the jaws

order to the Battle with the

1846, p. 153.
in the foregoing note.

of a sea-dragon, this

Amazons

in a series of the

deed standing next

in

Twelve Labors.

NimeBonner Jahrbuch,

Weizsiicker, in Roscher's Lexikon, Art. Hesione, refers to a relief at

guen, representing

Andromeda chamed,

vol. xxxiii. p. 66,

and

accessible to me.

On

vol. vii. p. 39,

as published in the

No.

6.

These volumes have not been

the other hand, both Weizsiicker and Baumeister have

failed to include in their lists the highly important relief for the
at Treves, representing the subject in a

manner

Arch

of Trajan

similar, yet decidedly, superior,

ARCHAEOLOGICAL LXSTITUTE.

236

Upon

sarcophagus and altar the form of the monster has


that of a dragon,

mained
Let
late

it

and

tail

re-

fin-like feet.

be observed that, without a single exception, these

works of ancient

art

all

that have hitherto been iden-

as bearing upon the archaic Greek story of Herakles

tified

and the sea monster


is

with curled

are

creations of the

Roman

Empire.

Homer

recounts but two of the exploits of Herakles

one

them.

of

It is

this

referred to by the great lyric poet of

by the Attic writers of tragedy and comedy, and by


prose authors of every Greek race and of every age yet not
a single work of Greek art has hitherto been admitted to
Boeotia,

illustrate

it.

Surely this

is

a fact not less remarkable than

the existence of so large a class of vase-paintings and reliefs

unconnected with any known myth.


If

ploit

now advanced

the conclusions

be

figures

drawn,

correctly

it

results

shown upon representations

must frequently be
been done.

respect to this ex-

in

that

of

subordinate

the

Herakles and Triton

otherv^ise explained

than

has hitherto

Thus, to give a single typical example, the

painting upon a vase formerly in the


collections, reproduced

the combat in a

on a small scale

manner

Durand and Pourtales


in

Figure

50, displays

fully characteristic of the large cate-

gory of black-figured vases.

We

see Herakles, clothed with

the lion's skin but wholly without weapons, bestriding the back
of the fish-tailed

monster, and holding him tightly around

the chest with hands interlocked in the labyrinthine grasp of

the palaistra.
to that of the

Three dolphins beneath show the struggle

Cologne sarcophagus, and published

in the

work mentioned

in a

preceding note.
Altogether uncertain

modern

is

the Trojan coin of Septimius Severus, retouched in

and thus rendered most untrustworthy, which Mionnet, Description de Medailles, vol. ii. p. 664, No. 224, describes as representing Herakles
crowned by Ilesione.
This completes the list of representations of this subject known to me.
times,

LWEST/GAT/OXS AT ASSOS,
Upon

to take place in the water.

the

left,

1S83.

237

Poseidon, accom-

panied by Amphitrite, hastens with uplifted trident to the


aid of his distressed emissary.
On the right, an interested
but powerless spectator, stands the white-haired Laomedon,
his royal state designated by the ruler's staff which he holds
in his hand,

by the coronet upon

FiG. 50.

The Struggle

his head,

and by

his wide-

of Herakles with Triton.

Painting upon a Black-figured Vase.

folding and sleeved mantle.

Behind him stands the coro-

neted princess Hesione, awaiting, with archaic impassiveness,


the issue of the conflict.
rf Tros, the

Beneath

this scene race the horses

promised reward of the

victor,

which play so

The appearance
upon other representations of the

]irominent a part in the subsequent story.


of these horses,

struggle,^

may

upon

this as

possibly serve as a further

argument

in fav^r

1 As, for instance, the Munich vase, described by Jahn, Beschreibimg der
Vasemammhinif, No. 391, and that published by Brondsted, Desci-iption of thirtytwo Ancient Greek Painted Vases, No. 7.

ARCH^OLOGICAL INSTITUTE.

238

of this identification of the exploit,

the only one with which

they can be connected.

In the

descriptions of this vase-painting, by

first

De

Witte^

and Dubois,^ the monster was identified as Nereus, the king


as Proteus, and Hesione and even Amphitrite as Nereids.
Nereus was correctly called Triton by Gerhard,^ as before
stated,

but

the

two figures upon the right were,

in

the

opinion of the present writer, wrongly designated as Nereus

Gerhard could not but

and Dons.

feel the difficulty of ex-

plaining the quiet spectator-like posture of this pair, which


contrasts so strikingly with the active advance of the other
sea gods upon the opposite side, and to justify this he

was

forced to the altogether baseless and unnatural assumption

combat between Herakles and Triton did not take


place until after Nereus had in some way been induced to
Although these subordinate
favor the interests of the hero.
figures seem so evidently intended for Laomedon and Hethat the

sione, this identification as

questioningly repeated by

Too
in this

large a

number

Nereus and Doris has been un-

modern

all

authorities.^

of vase-paintings

would require revision

sense to permit of a complete treatment of the sub-

ject in this place.

of Gerhard's

Suffice

it

to note that the very next plate

work^ shows the reward

of

the victor in the

quadrigas represented above, the fact of the struggle taking


place upon the coast being indicated
trees

and

in

on the isolated red-figured vase, so


Laomedon, characterized by the royal

that, finally,

frequently referred
staff

by dolphins and by

to,

and mantle, appears as a spectator of the struggle,

De

Dubois, Collections Poiiriales-Gorgier, No. 196.

Gerhard, Atiserlesene griechische Vasenbilder, vol. ii.


As, for instance, by Baumeister and Furtwangler.

Gerhard, Auserlescne griechische Vasenbilder,

De

Witte, Cabinet Dtirand, No. 302.

vol.

ii.

pi. cxi.

pi. c.xii.,

also described

Witte, Cabinet Durand, No. 304.

INVESTIGATIOXS
standinj;

before

palace,

his

AT

ASSOS,

1SS3.

239

by a single

naively indicated

column.

That upon the Greek representations


sione

is

of the struggle

never shown as chained to the rock, and

omitted altogether,

is

due

He-

generally

is

Greek legend

to the fact that the

does not describe her as present upon the shore on this occasion,

she having been previously released by Herakles, and

taken by him from Cape Agamias to Troy.


given by the Latin writers this

is

consequence invariably seen upon the

in

In the accounts

otherwise, and Hesione

Roman

Laomedon having defrauded Herakles

of the

is

works.

promised

re-

ward, the hero revenged himself, as well as Poseidon and


Apollo, by killing the deceitful

Troy, as mentioned

the

in

Iliad.^

monarch and demolishing


It

is

worthy

particularly

of note, as an evidence of the importance attached

to

this

legend by natives of the country during the historic period,


that Strabo^ found the inhabitants of Ilion offering no wor-

ship to Herakles, because, as they explained, of a feeling of

resentment which they entertained on account of


struction of their town.

Nothing could better

this

de-

illustrate that

antique spirit of local patriotism which leads us to attach


great weight to specifically localized traditions in
gesis of works of art such as this.

At Assos,

with

the exeits

Aeolic

population, Herakles was the object of peculiar veneration,

and

it

is

not strange that the one exploit of the national

hero which

was

intimately

connected

with

the

province

should have been represented upon the walls of the chief

temple of the Southern Troad.

The

defeat of the Triton

which threatened Hesione, and the consequent dethronement


of the unjust Trojan king, cannot but have been regarded as
1

///ad,

V. 640.

Compare

also the other authorities for the legend, quoted

above, especially Diodoros, IV. 49.


2 Strabo, XIII. p.
596.

ARCH^OLOGICAL INSTITUTE.

240

significant of the emancipation of the country


It

valor.

certainly presented

the decoration of the first


colonists of the Troad

by Hellenic

the most fitting subject for

monument

erected by the Greek

after their deliverance

from Persian

tyranny.

The choice had, moreover, a religious as well as a national


It was Athena who participated with the hero

relevancy.
in the

performance of

which was

by erecting the rampart

this exploit,

to serve for his protection

was upon the walls

it

of Athena's temple that the struggle was depicted.

The second
of the Louvre

considered

in size

and importance among the Assos

a corner block,

represents

four

like that

men

reliefs

which we have just

reclining at a symposion

They are waited upon by a fifth, who, standing


upright, is, by reason of the isocephalism, of much smaller
As the chief figures are more
proportions than the others.
inclined, the difference of scale is even more marked in this
composition than in that of Herakles and Triton.^ Each of
(Fig. 51).

the banqueters holds in his

most, at the

left,

which the attendant

of

left

hand

a drinking cup, the fore-

being provided with two such vessels, one


fills

from an oinochoe, replenished

from a huge krater standing behind him.

The only notable

act depicted

upon the

relief,

and that

which evidently forms the subject

of the scene,

sentation of a strap-like girdle by

the second figure

Texier {Description,

vol.

ii.)

remarks that the cupbearer

is

is

the pre-

on the

made smaller

be-

cause of his less dignity but this subordination is obviously the result of, rather
than the reason for, the great difference in proportions. In compositions such as
,

by so arranging the positions, while


adhering to the principle of isocephalism, as to give prominence to the chief figUpon Etruscan reliefs and wall paintings representing funeral banquets,
ures.
these, archaic artists displayed their ability

more

perfect scale of proportions

was rendered

possible, while keeping

all

the

heads upon the same level, by elevating the reclining figures upon couches.
Compare, for instance, the relief found at Chiusi, and published by Micali,

Monnmentiy

pi.

xxiii.

Even here the upright

figures are too small.

ARCH^OLOGICAL INSTITUTE.

242
right to the

heart in

The

first.

recipient places his

hand upon

his

an eloquent gesture of humble and almost hypocrit-

The scene

ical obligation.

thus depicted

is

conceived by the

present writer to be the delivery by Herakles to Eurystheus


of the girdle of Hippolyte, well

known

as the trophy of the

expedition against the Amazons.

be correct, the subject forms a contin-

If this identification

uation of the one represented upon the corresponding block


for

it

will

be borne

in

mind

that the contest of Herakles with

the sea monster, and the deliverance of Hesione, by which

the hero was so intimately connected with the Troad, was an

episode of that expedition which resulted

Amazon

in

the defeat of the

Apollodoros makes particular mention of

queen. ^

the fact that the girdle was given to Eurystheus by Herakles

And

himself.^

it

is

an accepted belief among

text, that, in his concise relation of

critics of the

the exploits, Apollodoros

has given us an abridgment of the work of Hellanikos of


Lesbos, greatest of the early logographers, who, as a native of
the parent city of Assos, writing
building of this temple

is

in

referable,

the very age to which the

would certainly be the best

possible authority for the version of the legend followed

the designer of these sculptures.


livery of the

the residence of Eurystheus was a

girdle in

detail of the story regarded with peculiar satisfaction

Greeks,

is

apparent from the words

ing this expedition

"

by

That, moreover, the de-

of

by the

Euripides concern-

For Hellas received the

rich spoils of

the barbarian maid, and they are safely kept at Mykenai."


1

Apollodoros,

p. 49,

II. 5. g.

It

was related by Hellanikos [Frag.

preserved by the Scholiast to Pindar, A^em.,

accompanied Herakles

in his

33, ed. Miiller,

III. 64) that the

Argonauts

expedition against the Amazons, and thus the ex-

came to be treated as a mere episode of this cruise. Compare


Diodoros, IV. 42, 49; Apollonios of Rhodes, II. 967 Valerius Flaccus, V. 132.
- Apollodoros, II. 5.
9 KOfxiaas de rhv (ajo-r^pa els MvKrtuas. eSooKfv EvpvffOfl.
8 Euripides, //ere. Fitr,,^\6
rot KXeiva 5' 'EAAots f\a$6 $ap0dpov Kopos \a(pvpa,
ploit gradually

Kal crw^er' iv MuK-fjuais.

INVESTIGATIONS
These direct references

AT

ASSOS,

1883.

243

may be

to the delivery of this prize

held to outweigh an isolated account of late date as to the


strained relations existing between Ilcrakles and lun-ystheus,

according to which the cowardly king, after the fright experienced

receiving the Erymanlhian boar,

in

occasion he had crept

away

on

hide himself in a brazen

to

which
jar,^

refused to give personal audience to the hero.^

The

feat of

Herakles

in

obtaining the girdle was one popu-

archaic art,^ and upon the few vase paintings which

lar in

display the girdle itself this object

Thus

shown upon our

that

ilar to

related by Diodoros, IV.

is

11

3.

of a form entirely sim-

With exception of the

relief.*

Apollodoros

(II.

5.

refers the

i)

Eurystheus to the sight of the Nemeaii lion. This picturesque story


of the king hiding away from the terrible beasts brought into his house is unfright of

doubtedly of great antiquity, being shown upon archaic vases referable to the
close of the si.xth century.

That the event was

not, however, originally, held


Herakles from the presence of his
royal cousin is evident from the fact that Apollodoros, in his subsequent account of the delivery of the Erymanthian boar (II. 5. 4) and of the mares of

to

have resulted

Diomedes

(II. 5

in the entire exclusion of

8), as well as of the girdle of

entering Mykenai, and on one such occasion,


bull, as himself

showing the animal

to

Hippolyte, describes the hero as

when returning with

Eurystheus

the Cretan

(II. 5. 7).

Diodoros gives a similar account of interviews between the two, and makes
no mention of any refusal to grant personal audience.
- Venetian Scholiast to the Iliad, XV.
639, following the passage of Apolloi), commented upon in the foregoing note.
The Victorian Scholiast
same passage of the Iliad gives another explanation of the relationship
between the hero and the king, which is in like manner recognizable as a perversion of the original legend, and well illustrates how freely such alterations
and additions were circulated during the later ages of antiquity. This is that
Eurystheus was the pederast of Herakles, who executed the labors at his behest

doros

(II. 5

to the

on account of this unnatural affection. If any weight at all be attached to these


embellishments of the tale, it must be admitted that the latter asserts the
continuation of personal intercourse denied by the former,
late

Compare

battente le
*

the review of this subject given by

Amazzoni, Annali,

vol. xxxvi.,

Roma,

Jahn

(Otto), Ercole com-

1S64.

As, for instance, upon the vase referable to the close of the

tury, published

ZeitJtng, Berlin, 1856, pi. 89.

fully given

girdle

is

ratiir

und

fifth

cen-

by Welcker, Herakles und die Amazonenkonigin, Archdologische

The

extensive ancient literature in regard to the

by Kliigmann (Adolf), Die Amazonen in der Attischen

Kiinst, Stuttgart, 1S75, particularly in notes 19 to 23.

The

Lite-

classic

ARCHAEOLOGICAL INSTITUTE.

244

girdle no attributes are represented,

and no peculiar features

distinguish the four reclining figures

put

provided

with a deep-lipped

others holding in the

The

type.

right
is

upon the

this interpretation

hand

left

unless, indeed,

drinking

we may

Herakles alone

fact that

bowl,^

hand a kantharos

each

had

it

the

of

of the ordinary

shallow vessel without handles, uplifted by the


of the foremost figure,

and

filled

by the attendant,

a phiale, such as was customarily used for libations

that

is

the fact that the

member

like the others, a

drinking vessel for his

of the

and

seems evident from

this case

this significance in

banquet who holds

own use

it

has,

in his left

hand.

An

two

identification of the

figures at the

left,

appear as guests at the royal table of Eurystheus,


requisite nor possible.

duced

for the sole

They belong

It is

purpose

who
is

thus

neither

obvious that they were intro-

of filling out the elongated panel.

to that class of figurants of

which the ancient

designers and vase painters possessed so large a retinue.

The

variety and elegance of the seven vases represented

upon the

relief

may, as

will

be argued

in a

subsequent con-

nection, be taken as an indication of the fact that the

work

is

not referable to any period more remote than the Lydian and
Persian invasion
and the same conclusion, if any, is
drawn from the reclining postures of the banqueters.
Greeks of the Homeric poems, as is well known, sat at
;

to

be

The
their

authors, however, devote their descriptions rather to the brilliancy and great

value of the trophy than to


1

It

may be observed

its

shape.

that Herakles holds a cup of the

same shape

in his

symposion with Pholos, as represented upon the archaic vase published by Gerhard, Herakles bei Pholos und Busiris, Archdologische Zettung, 1S65, and that
the
his position there is precisely that of Eurystheus upon the Assian relief,
The exleft elbow leaning upon a cushion, the right arm crossing the breast.
ceptional direction of the composition, from right to left, is also the same, and

likewise points to

some common prototype.

The

peculiar attitude

which Lucian {Conviv XIV.) refers as customary


banquet of Herakles and Pholos.

that to

is

doubtless

in paintings of the

INVESTIGATIONS
meals and drinking bouts
Sosias in the Berlin

;i

AT

ASSOS,

18S3.

245

and upon the celebrated vase

ot

a work of severe style, yet


Museum,
of Olympos are shown seated at their

red-figured, the gods

Reclining at table was originally an Oriental


usage as may be gathered from the domestic scenes depicted
upon Assyrian monuments, and it has been plausibly assumed

carousal.'-^

have been introduced to Hellenic life by the lonians.^ At


however, by
just what period this custom became general is,
quevi
can be
post
terminus
definite
no
and
certain,
no means
to

derived from the adoption of this posture upon the Assian


this,
Still it appears worthy of further consideration in
relief.
sense, that

riously

all

upon

the banqueters are here

shown

as leaning luxu-

cushions, in like Oriental fashion.

the Greeks had accustomed

Long

after

themselves to recline at table,

they continued to employ plain couches for this purpose.*


Even in the third century the Spartans " were wont to lie

upon bare benches during the whole banquet," and " hesi" which had come
tated to put their elbows upon the pillows
into fashion at court during the reigns of Areus and Akrotatos.^

Notwithstanding the obvious indication afforded by the


girdle, none of the writers who have discussed the subject of
XXIV. 475, 515 Odyssey, VII. 203, XXI. 89.
Furtwangler, Besch7'/;ibu>tg der Vasenscnnmhing im Antiqtiarhim, Berlin,
vase.
18S5, No. 2278, where see the very extensive literature concerning this
1

Iliad,

By

Miiller (Carl

Die Dorier, 2d

Otfried),

ed., Breslau, 1844,

Geschichten

IV.

3. i.

Helle7iischer

The author

Stdmme

itnd Stddte

further concludes, from

a passage of Alkman preserved by Athenaios (III. 75), that the Greeks of the
age of that poet reclined at meals, inasmuch as klinai are mentioned as being
provided for the guests. Alkman, however, who was himself a Lydian by birth,
appears to have described in these lines the lavishness of some Oriental banquet.
The similar change in the customs of the Romans was effected at a much later
Compare Servius, ad Aen.,
date, and was particularly mentioned by Varro.
VII. 176, and Isidorus, Orig.,

XX.

Ii. 9.

XVIII., Athenaios, XII. 15, p. S''"^; Souidas, s.


Cicero, Pro Miiraetta, XXXV.
*iA/tio and \\iKovpyos
5 Phylarchos, Hist., XV. and XX., preserved by Athenaios, IV. 20, p. 142.
*

Plutarch, Lyacrg.,

v.

ARCH^OLOGICAL INSTITUTE.

246

any way with the expedition


Herakles against the Amazon queen and the possession
have connected

this relief

of

Thus, Poujoulat conceived

misunderstood.

women upon

in

Indeed, none of the rehefs has been more

of this trophy.

"

it

it

represent

to

a couch, their long hair, which forms their

only covering, floating carelessly upon their shoulders."

The

block happens to be so fractured as to form two chief pieces,

each containing two banqueters, and

Louvre

walls of the

was attached

it

to the

Eu-

joint surface to joint surface, so that

rystheus rested his elbow against the krater at the other end
of the relief.

was thus displayed

It

to the public for half a

M. de

century, until, in 1886, the writer called the attention of


Villefosse to the matter,

This

adjoined.
rally

and the fragments were correctly

was thereby

(right half), and

who

more

still

Menelaos

two sepa-

and Proteus en pourparlers

Menelaos and Proteus d' accord

correctly

Clarac

unintelligible.

led to consider the relief as containing

rate representations,

Texier,

connect the figures natu-

failure rightly to

rendered the subject

combined the groups

in his

{\&ll half).^

engraving,^

described the scene as the feast of Perithoos, which, in view


of the

huge wine

jar,

and the centaurs

More modern

not so bad a guess.


1

Michaud

Lettre
2

et

Correspondance

Poujoulat,

critics

d' Orient,

and historians
vol.

iii.,

Paris,

of

1834,

LXIX.

Clarac, Music, vol.

ii.

The account is too good


" Nous retrouvons

seconde partie, Paris, 1S41.

not to be given as a last quotation from this delightful book

encore

was

of the other reliefs,

ici

Menelas

et

Protee

ils

sont a peu pres d'accord, et

sur la resistance opiniatre du dieu marin qui senible deja

le
lui

heros

I'a

emporte

avoir appris una

partie de ce qui I'interesse et qui, partageant avec lui la coupe de I'hospitalite,

cherche par ses demonstrations, et en posant sa main sur son cceur, a


vaincre de sa franchise.

Le Roi de Sparte

le

con-

n'a pas une entiere confiance en ses

protestations, et I'espece de bandelette qu'il presente d'un air serieux a Protee,


n'indiquerait-elle pas que,

s'il

ne

lui

tient

pas entierement ses promesses,

revenir encore a la force, et I'entourer de liens dont

de se degager
*

ume

il

ne

lui

sera pas

si

il

va

facile

"

Te.xier, Description

d^Asie Mineurc,

vol.

ii.

pi.

of L'L'ni-'crs entitled Asic Miticure, Paris, 1S62,

114; re-engraved in the volpi. 15.

INVESTIGATIONS
Greek

art

have been

concerning which,
tions

among

certainty

The

is

the

AT

ASSOS,

247

committal in regard to the subject,


the entire lack of parallel representa-

less

in

known works

of ancient art, an absolute

perhaps not attainable.

positions originally occupied by

tured

1883,

epistyle

blocks

are

ascertainable

many

of the sculp-

from an elaborate

having for its base the various widths of the


intercolumniations of fronts, sides, and corners, the
various
lengths of the regulas and half-regulas carved upon
the pancalculation,

and that peculiarity of the Greek Doric entablature


by
which the corner metope is removed from the axis of
the
corner column to the corner of the frieze.
To these definite
els,

there are to be added, as secondary indications,


the
kindred nature of the subjects represented in certain
cases
facts

upon adjoining

reliefs,

the direction of the compositions to-

wards the central panels of the fronts, and, finally, the


relative positions in which the overthrown blocks
obtained by
the

American excavations were

those

reliefs

French, the

discovered.

In the case of

which
last

were removed from the site by the


mentioned of these indications has, unfor-

tunately, not been put

on record. Moreover, these blocks


have been deprived of many of their characteristic
features,
such as pry-holes, corner joint surfaces, and
relative
thick-

nesses of boss and

soffit,

order to facilitate

their

by bemg sawed to thin slabs, in


attachment to the walls of the

Louvre.
In view of the shattered and defaced
condition of the great
majority of these stones, the results attamed
through this
examination cannot but be regarded as surprisingly

full.
Of
beams of the temple, only fifteen sculptured panels are now known and of
these fifteen, but a single
one remains entire. Only four are sufficiently
represented

the forty-four epistyle

ARCH^OLOGICAL INSTITUTE.

248
by fragments
lengths, and

to permit the

measurement

of those half-regulas

their original

relative position of every

may be

one or both

which alone can supply trustworthy

in-

Yet, notwithstanding this, the

dications as to width of span.

Thirteen

of

three of these, again, are deprived of

one of the

reliefs

is

now known.

assigned to their exact positions

and

in

regard to but two small fragments, belonging to the series of


wild beasts, can any doubt obtain as to the particular inter-

columniation which they occupied.

The lengths

of the entablature, side

and

front, not includ-

ing the projection of the tainia, or of the band bordering the


panels upon the lower edge,

may be

accurately ascertained

by subtracting from the corresponding dimensions

of

the

stylobate the lower diameter of the column, plus twice the

distance of the arris from the rise of the upper step, and

adding to

this result the thickness of the epistyle.

fronts the dimension thus obtained


for the

is

13.89 m.

For the

Assuming,

purpose of preliminary examination, the columns of

the facades to have been equally spaced,

it is

evident that the

corner panels would have a length of about 3.03 m., while

beams would have averaged 2.61 m. The


length of the side entablature is in like manner found to have
been 30.17 m. As it is known from the marks upon the
the three inner

stylobate that the corner intercolumniations of the sides were

somewhat

larger than the others,^

the shafts being in one instance

the clear opening between


the corner blocks
1.568
m.,

must have been 2.9 m. long, and the others have averaged
We might hence expect to find four distinct classes
2.44 m.
of epistyle beams, respectively 2.44, 2.61, 2.90, and 3.03 m. in
length.
It is to

be borne in mind, in this examination, that the

half-

regulas carved upon the ends of the panels are often consid1

See page 76 of the present volume.

AT

INVESTIGATIONS
erably

longer or

ASSOS,

1SS3.

249

than one half the width of the

shorter

triglyphs, thus proving the joints of the lintels to have

displaced from the axes of the columns.

been

While the triglyphs

average 52 cm. upon the sides, and 56 cm. upon the fronts,
the half-regulas vary from 11 to 38 cm., showing the joints to

have been occasionally as much as

15

cm. out of centre.^

In

the calculation of the corresponding intercolumniations, the


plus or minus, thus definitely ascertainable, has of course to

be taken into account.

The

be recorded

first fact to

style blocks discovered

is,

upon the

that the unsculptured epi-

site were,

without exception,

of a length corresponding to an intercolumniation not greater

than 2.45 m., and are consequently to be assigned to the


sides of the building.

Turning
corner

lap,

to the reliefs,^

we

must have had a

find three which, including the

total length of

over three meters,

and consequently must have occupied three


ners of the facades.

of the four cor-

These are the Herakles and Triton, the

Banquet, and the Herakles and Pholos.

further proof of this fact, derived from differences in the character of

the tooling upon the


the present
2

soffits

of the epistyle beams, has

been given on page 85 of

volume

The twenty-two drawings

of the

Assos

reliefs

following in the text are

given, not as adequate illustrations of the sculptures, but as displaying

blocks in their character as architectural


lintel

when complete

is

members

The

all

these

each
given in centimeters by the figures beneath, those above
total length of

indicating the length of the regulas, and of the spaces betwen them, which correspond to the metopes.
The writer is responsible for these measurements,

which were taken

and

Assos from the blocks themselves. In the


Louvre he has, however,
followed a series of sketches made by Mr. Bacon prior to the commencement of
the excavations
These sketches were not drawn to strict scale, and those who
desire to verify the arguments dependent upon the dimensions of these reliefs
should base their calculations upon the figures accompanying these cuts and
given in the text. The dots follow the architectural lines obliterated from the
blocks, and give the probable length of the missing members. They furthermore
show the deviation of the lintels from the normal lengths by indicating the axes
in Paris

at

outlines of the sculptures which are preserved in the

of the columns.

archjOlocical institute.

250
The

first

served in
2.95

mentioned (Fig. 52), though fractured, is preentire length, which measures not less than

its

The

m.

half-regula at the left-hand side of the block

about 8 cm. too long, and this amount

is

to

is

be deducted in

calculating the width of the corresponding intercolumniation.


But, as an entire

regula

20 cm. must be added


lintel

from axis

in

is

not cut at either end, at least

order to

and

to angle;

make up

the length of the

this total of 3.05 furnishes a

decisive proof that the stone was above one of the corners of

Fig. 52.

Epistyle Block above the northernmost Intercolumniation


OF the Eastern Facade.
Series relating to the Rescue of Hesione.

the fronts, and not one of the corners of the sides.

It is fur-

thermore evident that the laps forming the corners of the


entablature were cut upon the epistyle blocks of the sides,

and not upon those of the

fronts.

Ample grounds

for the

adoption of this arrangement are to be found in the consideration that

the

and the builders

quarrymen were thus required


to handle, eight blocks of but 2

to provide,

90 or

2.95,

instead of four of 2.70, and four of not less than 3.10 or


3.15 m.,

an immense

the primitive builders,

saving in practical respects, of which


ill

provided with machines for trans-

porting and lifting such heavy stones, must have well been

aware.

Even

as

it is,

the heaviest and most

these front corners of the epistyle are

awkward stones employed

in the edi-

AT ASS OS,

INVESTIGATIOXS
fice,

1883.

25

weighing half as much again as the bulky corner cornice

blocks,

another

and being much more


relief is

known

to

of the eastern front, while

difficult to

The

set.

fact that

have occupied the southern corner

two

an entirely different

reliefs of

subject adjoined the northern corner block of the western


front upon either hand, and, above

ment of the composition,


hand end

leave

all,

and not the

of this panel,

the grouping and move-

doubt that the right-

little

hero has driven the monster into a corner


spectators

The

was outermost.

left,

the affrighted

From

towards the middle of the entablature.

fly

a decorative point of view, also, the broad masses formed by


the bodies of the combatants are of decidedly better effect at
the outer end, the upright lines of the smaller figures at the

Hence

inner.

right

hand

the lap has been

of the lintel

drawn

Figure 52 upon the

in

and we only remain

in

doubt whether

to assign the block to the northern corner of the eastern, or

the southern corner of the western facade,

gard to which no immediate decision

were found.
of

the

point in re-

possible, as

is

no information concerning the position

in

we have

which these

reliefs

Certain indications derived from the spacing

metopes, and

pointing to the

probability that

the

Herakles and Triton occupied the eastern, and the Banquet


the western fagade, will be adduced in another connection.

The

distance

from

the

central

regula

to

the joint

sur-

upon the right hand of the lintel is 1.12 m. At this


end the half-regula has been split away, but it appears
face

scarcely possible that the

length than 27 cm.

one instance

As

member can have had


the corner triglyph

is

a greater

known,

in

have been 52 cm. square in plan, it


follows that 25 cm., or thereabouts, is to be added to the
at least, to

actual length of the lintel.

Thus

it is

evident that the entire

panel, from joint to angle, cannot have been less than 3.2 m.
in lencrth.

ARCH^OLOGICAL

252
It

the

was more

relief of

obtain an accurate measurement of

the Banquet, this being broken into four pieces,

which were,
walls of

difficult to

IiXSTITUTE.

at the time of examination, arranged

the Louvre

in

together.

closely fitted

incorrect sequence, and


Still,

upon the
not very

the total length of 2.86

given upon the drawing (Fig.

53,

compare

hardly involve a greater error than

Fig.

51),

The

one inch.

m.,

can
half-

regulas in this case are both excessive, measuring 36 and

38 cm.

The

larger of these

was outermost, and adjoined

Epistyle Block above the southernmost Intercolumniation


OF the Western Facade.

Fig. 53.

Series relating to the Girdle of Hippolyte.

the lap, the thickness of which hence cannot have exceeded

may

not have been greater than

14 cm.,

18 cm., while

it

being

case the thinnest block of andesite employed

in either

in the construction of the temple.

stricting the width of the lap

The

object of thus re-

was evidently

to

extend the

sculptured surface of this panel as nearly as possible to the


corner.

The

length of the epistyle from axis to angle was

about 2.96 m.
3.04 m.
left,

of

the entire panel,

That the right-hand end

adjoined the corner

is

about

including lap,

of the block,

and not the

indicated by the direction of the

the banqueters naturally facing the


the fagade, as well as by the points referred

composition,

middle of

to in connec-

tion with the relief of Herakles

and Triton, which

is

thus

INVESTIGATIONS

AT

ASSOS,

1SS3.

253

seen to have been placed upon the corner of the building


diagonally opposite.

The

third front corner block

^
that sculptured with the

and Pholos, discovered during the Amer-

figures of Herakles

(Fig. 54.)

ican excavations.

is

It

presents fewer factors for

the calculation than do the others, inasmuch as the two frag-

ments remaining do not constitute the entire lintel, a considerable portion, including one of the half-regulas, being
missinof from

FiG

54.

the

left

hand

Furthermore, the

side.

half-

Epistyle Block above the southernmost Intercolumniation


OF the Eastern Facade.
Series relating to the Centaurs of

Mount Pholoe.

regula from the other end has been split away from the sur-

From

face.

the right

the end of the central regula to the joint upon


1.05

is

m.

and

the fracture upon the


half-regula

is

83 cm. from the other end to

is

it

left,

where no commencement

although the tainia

visible,

Measured along the lower edge


recovered

is

2.43

m.

and

if

of the

perfectly sharp.

of the tainia, the total length

the dimension of the missing

member be added to this we have


much too long to permit
2.7 m.,

the inner intercolumniations.

That

a front, and not a side corner,

is

half,

is

a lintel of not less than


us to assign
it

it

to

any

of

furthermore occupied

evident from the fact that a

and not a whole regula, was cut upon the end remaining

intact,

which, as

may be concluded from

the direction of the

composition, and from other indications, would be that occu-

ARCHAEOLOGICAL INSTITUTE.

254

pying the angle of the building,

if

the relief had been assigned

All these considerations are in agreement

to a side corner.

with the theory that this relief was placed at the southern
corner of the eastern front, beneath which

it

By

was found.

adding to the extent of the remaining fragment the dimension


of the corner regula,
to angle,

we

see that the entire panel, from joint

must have been

at least 2.95

m.

in length,

and

may

have been more.


It will

subsequently be shown, by a calculation based upon

the lengths of the two adjoining blocks, that, assuming the

middle triglyph of the front to have occupied the exact centre


of the entablature, this corner stone, plus the width of the

angle lap, must have had a length of exactly 3.085 m.

or, in

other words, that the panel from axis to angle was precisely
the 3.03 m. requisite according to the width of the correspond-

ing intercolumniation.
a fragment of

It

is

thus susceptible of proof, that

the relief has been broken

away

enough to contain the equine body of Pholos,


which reference has been made

in

just large
a

fact

to

the consideration of the

sculptured subject.^

similar calculation,

symmetry

if

it

be not affected by a want of

in the position of the central triglyph of the front,^

proves that the disproportion between the widths of the two

metopes above the

relief

even exceeded the very considerable

Page 151 of the present volume.


The probability that some slight correction

in this

sense

order to determine the exact dimensions of these epistyle blocks,

is
is

necessary in
indicated by

the fact that the stele upon which the confronting sphinxes of the eastern facade

and which marks the centre of the symmetrical composition, is


removed about two centimeters from the axis of the regula above it. It is,

rest their paws,


itself

however, impossible to make allowance

for a variation of this nature, in the ab-

sence of one of the epistyle blocks of the front.


text

must approximate very

The dimensions given

closely to the actual sizes,

in the

and serve the most impor-

tant purpose of this consideration, namely, the determination of the positions of

the

known

reliefs,

and the projected widths of the intercolumniations.

IXVESTIGATIONS

AT

ASSOS,

1SS3.

255

amount already evident from the spacing of the rcgulas upon


The metope on the right hand can-

the fragmentary block.

not have been btoader than 74 cm.

that

upon the

left is

seen

from the space remaining upon the relief to have exceeded


83 cm., and is now shown to have been not less than o.i m.

more than
m.

0.2

little

The

this.

at first sight

is

astounding

yet

it

is

relief itself,

proportionally

cm. respectively.

In the

and thus admit

of

no possible doubt.

Moreover,

ascertainable, even from the few metope slabs found

is

upon the
than

site,

that these

members

varied

still

more

in

width

indicated by this calculation, namely, from 6^ cm. to

is

mm.

An

explanation of such inequalities in the division of the

frieze, particularly

suggest

by the

itself in

builders.

above these two front epistyle beams,

It

has been seen, from marks upon the sty-

which form the steps

was commenced near the northwestern corner


and was continued
southern

lowed

in the

and

in

to

construction of the entablature.

frieze

members

of the eastern front

direction from north to south.

been carried on

in

of the edifice,

both directions until they met upon the

The same sequence appears

side.^

will

reviewing the method of construction adopted

lobate, that the laying of the stones

tels

nearly

the unequal dimensions can be taken from the

latter case,

905

of

more than that between the metopes above the adjoin-

ing block, measuring 68 and 81

it

between them

difference

have been

fol-

Thus, the

lin-

were

The work having

laid in the

evidently

the most primitive and irregular manner,

without the aid of scaled working drawings, or accurately de-

termined tables of dimensions, everything had to be done by


1

Page 64 of the present volume.

The

considerations which determined this

course of construction were undoubtedly connected with the facts that the stone

was brought to the site of the temple from the northwest, there being no approach to the Acropolis from the south, and that the native rock reached the
highest level at this part of the plan.

'

INVESTIGATIONS AT ASSOS,
testing and fitting.

The temple was

1SS3.

257

same way

built in the

as

were the polygonal city walls of that epoch, with their irregular angles and unequal lengths of the separate stones.
For
width of every second

the

measurements had
of construction.

to

metope, at

least,

independent

be taken from the edifice

in the course

The want

of agreement between the joint

surfaces of the lintels and the axes of the columns

system, and

result of this

we have been

was one

forced to conclude

that the exact lengths of the regulas were cut

upon the

face

in position

members of the frieze had been placed


above them. The only fixed principle in the ar-

rangement

of

of the reliefs after the

the frieze was that every alternate triglyph

should be placed as nearly as possible

umn

beneath

it.

Thus

it

in

the line of the col-

was brought about

that the entire

correction rendered necessary by the elongation of the

mem-

bers of the frieze to correspond with the longer

and especially

fronts,

second metope

laid

the corners,

of

on each

lintel,

under consideration that upon the


This

beams of the
was thrown upon the

this

left

being

in

the case

hand.

be rendered clear by retracing the steps naturally


taken by the masons. As shown in Figure 55, which o-ives
will

in isometrical projection the dimensions of the southeastern


corner of the building, the triglyph A' was placed as nearly

as possible in the axis

of the

jointing of the lintels beneath

column A, regardless of the

To

it.

this triglyph adjoined

the metope A'\ for which no width could be directly measured upon the epistyle, and which was consequently cut of
the

mean dimensions, about

hand

0.7 m.

those upon the right-

sides of the reliefs of Herakles

and Pholos, and the four


horse-legged centaurs being but about an inch longer and an
inch shorter, respectively, than the average of the side metopes.

Then

direct

measurements were

followed the triglyph

B',

available,
17

for

which likewise no

and which was made of

ARCH^OLOGICAL INSTITUTE.

258

the width of 56 cm. determined for these members.


firmation of this,

it

may be

entire regulas appearing

upon the facade

The second

from the average.

tion of adjustment could

member, the

B'',

Any

C^

an inch

determined the position of

perceptible deviation of this

ideal continuation of the line of support in the

entablature, from the axis of the


intolerable.

reliefs varies

triglyph being set, the ques-

no longer be ignored, inasmuch as

the width of the next metope,


the columnar triglyph

In con-

observed that not one of the inner

Hence

column C, would have been

the location of C, thus fixed, was marked

upon the upper surface of the


deduced therefrom.

It

is

epistyle,

and the width of B"

obvious that so ill-considered a

manner of adjustment inevitably resulted in differences beAnd it is entirely in accord with this
tween B" and A".
explanation that the

maximum

irregularity should occur be-

tween the metopes above the front corner block, Q" and D",

upon which devolved the equalization

of the displacement of

the corner triglyph E' from the axis of the corner column

by not less than 15 cm., or half the thickness of the entablature minus half the width of the corner triglyph, as seen from
the calculation of the tainia and regula lengths of the relief
of

Herakles and Pholos, Figure 54.


This observation concerning the two metopes situated

above one and the same elongated epistyle beam, according


to

which that

last

placed in position tends to become the

broader, has direct bearing upon the assignment of the corner

blocks of the Banquet and the Herakles and Triton to the

Of

eastern or western facades.

these

reliefs,

that represent-

ing the marine monster has been seen to be the more important, in respect both to

Greeks

of

Assos and

to

its

its

which the temple was dedicated.


rial

national significance for the

connection with that deity to


It is

also superior in picto-

treatment, while in decorative composition

it

balances the

AT

INVESTIGATIONS

ASSOS,

1SS3.

259

Herakles and Pholos, which occupies the southern


corner of the front, far better than would the Banquet. The
six figures of the affrighted spectators who fly from the scene
relief of

of the struggle correspond

well to the discomfited

treating centaurs of the pendant.

might have

and reThese considerations alone

sufficed to justify the relegation of the

the western fagade.

That the arrangement

is

Banquet

correct

is

to

fur-

ther indicated by the spacing of the regulas upon the epistyle


in question.
The construction of the frieze having,

beams

according to our hypothesis, been carried on above the Herakles and Triton from right to

from

to

left

right,

left,

and above the Banquet

we should expect

to

find

the left-hand

metope above the former and the right-hand metope above


the latter to be larger than their fellows.
the case, by the amounts of 4 cm. and

This

cm.

is

An

actually

exchange

of the positions of the reliefs

of a double exception

would involve the assumption


to the natural law which they thus

exemplify.

We may

deduce the width of the corner intercolumniations


from the data already acquired. If from the

of the facades

smallest possible length of the Herakles and Triton panel,

namely, 3.15 m., we subtract 8 cm. for the excess of the halfregula at the inner end, and 42 cm. for half the thickness of
the entablature, we shall have a spacmg of about 2.65 m. on
centres

and

from joint

if

from the

total

length of the Banquet panel

we make a similar deduction, we have


about 2.54 m.
The difference between these results is to
be ascribed to the irregularities in the spacing of the frieze
members. It is to be borne in mind, that the position of the
to

angle

triglyphs in respect to the axes of the columns often


varied

even more than


visible

upon the
1

this, as
soffits of

has been seen from the bed toolings


the epistyle beams.^

Compare page 85

of the present

volume.

Now

the cor-

ARCHAEOLOGICAL INSTITUTE.

26o

ner columns of the fronts are known, from traces upon the

have been 13.07 m. distant from centre

stylobate, to
if

the shafts between

lar intervals,

This

is

of the

them had been placed

to centre;

at perfectly regu-

we should have an intercolumniation

of 2.614

i^-

very nearly the distance represented by the average

two corner beams preserved

the one of which

from axis

is,

the other two inches longer

to axis,
;

in

their entire

length,

but two inches shorter,

and with

dimension the

this

length of the relief of Herakles and Pholos, determined by


a

method

entirely independent,

is

in perfect

Had

agreement.

the temple followed the customary proportions of the Doric


plan in this particular, a very considerable difference between

Thus

the intervals, say 0.2 m., would have been observable.

there can be no doubt concerning the fact that the design


called for an equal spacing of the

columns

trary to the general usage of the style.

refinement

is

entirely in accord with

The

of the edifice.

inequality of

in the division of the frieze,

was so

its

of the fronts, con-

The want

of this

the general character

dimensions, especially

great, that the aid in the

adjustment of the corner metope, commonly derived from a


diminution of the corner intercolumniations, was not

be requisite.

felt

to

This neglect of one of the most characteristic

features of the style

means an evidence

is,

as will subsequently be

shown, by no

of great age, but merely the result of pro-

vincial rudeness.

The

equal spacing of

the intercolumniations being thus

determined, we can proceed with greater certainty to a consideration of the other epistyle blocks of the fronts.
readily recognizable

sphinxes.

That assigned,

western front,

The block

among

is

is

Most

these are the reliefs of heraldic

for reasons already

preserved in

its

adduced, to the

entire length.

broken into three fragments, two

(Fig.

of Avhich

56.)

were

found during the American excavations, while the third

is

IXVESTIGATIONS
among

AT

ASSOS,

1SS3.

261

the reliefs of the Louvre, so that

it was not possible to


the fractures together for the purpose of accurate
measurement. The given total of 2.58 m. cannot, however, vary
more
than two centimeters from the original length.
Fortunately

fit

Fig. 56.

Epistyle Block above the Central Intercolumniation


OF THE Western FAgADE.
Coat of Arms of Assos.

both half-regulas are preserved, and these, as has been


seen,

have sufficed to determine


appertained.

As

these

to

which

members

of the facades this relief

are short, about 8 cm.

be added to the length of the stone

Fig. 57.

in

is

to

order to obtain the

Epistyle Block above the Central Intercolumniation


OF the Eastern Facade.
Coat of Arms of Assos.

width of the corresponding intercolumniation,


which is thus
seen to have been some 5 cm. longer than
the average,

as-

suming, for the purpose of this calculation,


the axes of' columns and triglyphs to be exactly identical. The relief
from
the eastern front (Fig. 57) is lacking a
considerable
portion.

ARCH^OLOCICAL INSTITUTE.

262

including the right half-regula, while the

been

split

away from the

surface.

It

is

half-regula has

left

nevertheless ascer-

from the space remaining upon the stone and

tainable, both

from the length of the complementary half-regula upon the


adjoining block, that the

The

23 cm. in length.

distance from the remaining joint surface of the lintel

to the centre of the

composition,
5

member was about

middle regula, and of the symmetrical

By adding

1.26 m.

is

supplementary

to this a

cm. for the short half-regula, and doubling the

result,

have 2.62 for the width of the central intercolumniation.


is

thus proved that both these reliefs of sphinxes

vi^ere

we
It

above

inner intercolumniations of the fronts, and that these can

have been no others than the central openings

is

sufficiently

obvious from the absolute symmetry and heraldic character of

So

the figures upon them.

doubt

little

to this point, that the sphinxes

possible in regard

is

were assigned

to this position

even by Texier's restoration of the facade of the temple, upon

which every other


It is

worthy

relief is

wrongly placed.^

of note, that the three half-regulas

be measured upon the two sphinx


deficiency amounting

in

which may

reliefs are all too short,

one case

the

The blocks in
been made smaller

to 7 cm.

question are thus seen to have intentionally

than those adjoining them, and shorter than the intercolumniations over

An

which they were placed.

lies

near at hand.

two

figures,

The

explanation of this

sculptured subject, consisting only of

was without doubt

felt

events proportionally too high, to

to be too small, or at

fill

all

a panel of the normal

While this difficulty was met, in so


was possible by such adjustment, a corresponding advantage was gained by transferring the deducted length to

length and proportions.


far as

Texier, Descriptio?t,\o\.

relief of the

ii.

pi.

112.

The author

especially refers to the

Banquet, one of the two largest blocks removed to Paris, as having

been upon the side of the building.

lA'VESTIGATIOXS

AT

ASSOS,

263

1SS3.

the neighboring epistyle blocks, containing a greater

Of the two sphinx

of figures.

reliefs,

that

number

upon the eastern

was the more shortened in this manner, the half-panel


measuring 1.26 m. and it has been pointed out, in considering the style of these sculptures, of what decided advantage

front

even

to the design

As

still

this slight

further,

if

slight,

diminution proved to be.


indication of the correctness

of the assignment of the Herakles and Triton to the eastern


facade,

Fig. 58.

it

may be observed

that the longer corner relief thus

Epistyle Block above the second Intercolumniation from


THE south of the EASTERN Fa^ADE
Series relating to the Centaurs of

Mount Pholoe.

becomes complementary to the shorter central beam, and


z'crsa, excess and deficiency corresponding very closely.

vice

Fortunately for our understanding of the arrangement of


the reliefs upon the building, one of the sculptured blocks
discov^ered during the

American excavations

fills

the gap be-

tween the Herakles and Pholos and the eastern sphinxes, having formed the

main

lintel of

the second intercolumniation of the

front from the southeastern corner.

This

is

the relief of

the four horse-legged centaurs fleeing from the arrows of the

hero (Fig.

with

all

belonged
once.

58).

its

to

The

stone, preserved in its entire length,

mouldings

intact, is

2.6

m.

in

length.

and

That

it

an inner intercolumniation was thus evident at

Despite the striking difference presented by the horse-

ARCH^OLOGICAL INSTITUTE.

264

legged and human-legged monsters of the two blocks, it was


further assumed that it adjoined the Herakles and Pholos
group, next to which

it

at the southeastern corner

was found,

beneath its original position. And this


assumption has been confirmed through a comparison of the
half-regulas carved upon its ends with the corresponding
of the building, just

members upon the blocks on

That on the

either hand.

right,

measuring 33 cm., is exceptionally long, in agreement with


the small space remaining for the other half upon the intenthat on the left,
tionally shortened sphinx block (Fig. 57)
;

measuring 24 cm.,

in like

manner corresponds

to the space

upon the right-hand end of the Herakles and Pholos


Still

another proof of contiguity

supplied by the total

is

length of the half front entablature


three

reliefs,

ation

is

to

centaurs,

represented by these

accurate test, inasmuch as no great vari-

here conceivable.

panel, 3.085

added

an

relief.

The

total

length of the corner

joint to angle, as before determined,

m. from

the 2.6 m. of this relief of the four horse-legged

and

to the

1.26

m. represented by the half-panel

of the eastern sphinxes, gives just that total of 6.945

m.

re-

required for the half-entablature length, definitely determined

by measurements

of the plan.

which are

to

be assigned to the southeastern corner of the building,

we

Continuing the examination of those

have next

to deal with

two epistyle beams

resenting centaurs (Figs. 59 and 60).


in the

same

reliefs

in the

Louvre rep-

Their regular gallop

direction, their conventional positions, so similar

to those of their brethren

upon the

rally lead to the supposition that

reliefs of

the front, natu-

they were connected with

the exploit of Herakles against these monsters, of which the


chief scene is depicted

upon the front corner

the line of retreating centaurs

is

known

to

and

that, as

have been termi-

nated by the heraldic sphinxes occupying the middle of the

INVESTIGATIONS AT ASSOS,

1SS3.

265

facade, these wine-attracted disturbers of the hospitality of

Pholos are hastening to the assault, around the corner, upon


the side entablature.

It

shown

the sculptor has

may

all

furthermore be observed, that

but one of the centaurs retreating

discomfited before the arrows of the hero to have lost their

weapons, while

weapons

all

but one of those advancing have their

hands. Convincing indications of the


correctness of this view of the incident, and of the arrange-

ment

in their

still

of the blocks in accordance therewith,

Fig. 59.

derived

Epistyle Block above the easternmost IntercolumNIATION OF the SOUTHERN SiDE.


Series relating to the Centaurs of

Mount Pholoe.

from the dimensions of the two Louvre


eration.

may be

The

first

(Fig. 59)

is

reliefs

under consid-

recognizable as having belonged

to a side corner by the great length of its central regula, and


by the distance of this member from the remaining joint surface.
While the inner side lintels averaged, as has been
shown, but 2.44 m. in length from axis to axis, the corners

were not

less than 2.9

nation between them

From

m. long from axis to angle.

is

the ideal axis, 15

Discrimi-

hence neither dif^cult nor uncertain.


cm. beyond the left joint surface of

our block, to the centre of the middle regula,

is 1.32 m.; twice


plus the length of the half corner regula which is to be
included in the calculation, gives precisely the required di-

this,

mensions.
axis, that

It

will

be noticed

in

connection with this ideal

the stone presents a further peculiarity indicative

ARCH^OLOGICAL INSTITUTE.

266
of

its

above one of the corner intercolumniations

position

the half-regula at the

left is

the shortest

member

cut upon any of the epistyle blocks known, being only


in length.

It is

evident that the

beam

kind

of the
1 1

cm.

thus shortened

viras

through the same desire to spare the labor of quarrymen

and masons which determined the cutting

of the laps

upon

these corner blocks of the sides, rather than upon those of

the fronts.

It

was much easier

and handle two

to provide

blocks of 2.6 and 2.75, than two of 2.45 and 2.9 m., respect-

The

ively.

siderably

length of the whole regula

is

over 57 cm., con-

more than the ordinary width of the

side triglyphs,

For a corner

have averaged about 52 cm.

which appear

to

block this

entirely normal, such an extension being the

is

most natural method of effecting an adjustment between the


For the same
different lengths of the two classes of lintels.
reason, the metope

was allowed a width of 76 cm., while the

average width of the metopes above the inner intercolumniations of the sides cannot have exceeded 71 cm.
indication,

adduced

to

Still

another

and one of an entirely different nature, may be


prove the position of the block.

If

the panel ex-

tended to the corner of the entablature, with an entire regula


end, as indicated by the dotted lines in

at the right-hand

Figure

59, just sufficient

space would have been provided to

contain the figures of two centaurs in advance of the fore-

most now remaining.


case had the

lintel

This would by no means have been the

been above an inner intercolumniation, and

terminated by a joint surface in or near the axis of the next


regula upon the right.

centaur upon the

left is

It

is

true that the body of the last

divided by the joint

but this single

instance of the kind in any of the reliefs was due, as will presently be explained, to this very elongation of the corner block

by a

lap,

owing

to

which nine centaurs could be represented

upon two adjoining panels, instead

of four

upon each.

AT

IXVESTIGATIONS
The second

ASSOS,

of the Paris centaur reliefs (Fig. 60)

small dimensions which prove

it

to

ula.

is

cm. beyond the point

The

average of

latter
its

moulding

only

267
is

of the

have been above one of

the inner intcrcolumniations of the sides.


half-panel remaining intact
axis, 2

1883.

1.2

The

length of the

m. from the right-hand

to the centre of the

middle

ref^-

considerably shorter than the


class, measuring but 48 cm. in length.
This
is

block obviously did not adjoin the corner hntel the missing
hind quarters of the last centaur upon that relief are
not
;

Fig. 60.

Epistyle Block above the second Intercolumniation


FROM THE east OK THE SOUTHERN SiDE.
Series relating to the Centaurs of

sculptured upon

Mount Pholoe.

while the half-regulas are by no means


complementary. It is to be assigned, without doubt, to the
third intercolumniation of the southern side, from
the east.
If

it,

we assume

the joint surface upon the left-hand side, now


have been displaced a few centimeters beyond the
ideal axis, as would naturally have been the
case in the first
sculptured lintel, ample space would have been
provided for
the body of a fourth centaur.
Even had joint and axis
lost,

to

co-

incided,

this

figure

might have found

room.

It appears
probable that this fourth centaur terminated the long
line

of

advancing assailants, and that the adjoining epistyle


block
left was unsculptured
yet this point cannot be

upon^ the

definitely determined.

The arrangement

of the reliefs

upon the

six epistyle

beams

fj

z
^
o
C/J

M
Q
-

C
O
X
i-l

(In

2
O
o
n
tyj

td
Z
cs

<

u
f

y
T"
r-

<-"

<

u
S
H

(6

(X.

O
z
o
h
Pi

5^

tn
Ui

u
n
J
P3

5
a,

<

o
O

5 w

c ^
< hJ
< z

X
H
D
O Z
D
O
l

tc

O rn
ai
z S
o <!
2
r)
^ M
Oi
y.

Pi

U
W

ARCH^OLOGICAL INSTITUTE.

270

shown
Observe the difference between the

nearest to the southeastern corner, thus ascertained,


in Figures 61

and 62}

easy canter of the centaurs advancing


attack,

and the headlong

is

in regular file to the

who

flight of those

before the victorious arms of the hero

retreat in terror

the node of the ac-

tion, so to speak,

being the comparative repose of the upright

figures of P'holos

and

his guest,

emphasizing the vertical lines

the corner of the entablature in excellent architectural

at

effect.

One

other epistyle relief can be proved to have appertained

to a front of the building;

discovered during the

namely, that of the lion and boar,


year of the American excavations.

first

The fragment, comprising about

(Fig. 64.)

three quarters of

the entire panel, measures 1.34 m. from the remaining joint


surface to the middle of the central regula.
^

The

restored

elevation

of

the

gable

Deducting 4 cm.

and gable ornaments, given

Figure 62, introduces architectural features which require some words of

in
fur-

ther explanation.

The

tympanon

three stones of the

wall veil, referred to in the description

of the edifice, pages 106-109 of the present volume, are indicated by dots.
is

It

not absolutely certain that the blocks recovered appertained to the eastern,

and not the western gable

yet the position in which they were found favors

the former attribution, while the fact that the three stones belonged together

and

originally adjoined

is

evident from the exact agreement in height of their

corresponding sides.

The

relief

decoration of the terra-cotta gutter

is

based upon the indications

afforded by the fragment discovered during the second year of the excavations.

(Compare page

133, Fig. 31.)

The

height of this

member

is

the chief feature

which remains uncertainIn respect to the

griffin

drawn as the corner acroterion,

it

is

obvious that a

paw (page 137, Fig. 34) is by no means sufficient to warrant a trustworthy


restoration.
The figure is introduced merely for the purpose of indicating the

single

ascertained fact that a monster of this nature, whether griffin or sphinx, occuIts height, which may appear excessive, is but one meter,
and has been made proportionate to the dimensions of the acroterion surmounting the apex of the gable.
This central acroterion, dkoupe from a slab eight centimeters in thickness,
may be reconstructed with reasonable certainty as to its main features from the

pied this position.

existing fragment of its left-hand lower convolution.

(Page 136, Fig.

33.)

The

AT ASSOS,

INVESTIGATIOiXS

18S3.

271

for the excess of the half-regula,

the lintel

is

and doubling the remainder,


found to have corresponded exactly to an inter-

columniation of 2.6 m., or one of the inner spaces of the

Fig. 64.

Epistyle Block above the second Intercolumxiation


FROM the north OF THE WESTERN Fa^ADE
Series relating to the

It is plain

fronts.

that

it

Erymanthian Boar.

cannot have occupied a corner,

for the direction of the composition


fragment

in question, indicated

shows that a longer

half-

by sliading,

is shown, together with the outline


larger scale in Figure 63. As the volute
sends off a main branch upon the side opposite to the incised parallel lines
which evidently designate a horizontal juncture,

of the suggested reconstruction,

it is

plain that the

upon a

ornament must have consisted

of at least two pairs of scrolls.

remaining

is

The fragment

not less than 55 cm. broad, showing

the width of the scrolls to which

it belonged to
have been very nearly one meter. The given
dimensions of the acroterion are thus by no means
too large. At first sight, so broad and bulky
a mass will appear disproportionate to the gable.
But it is to be borne in mind that the temple of
Assos was at once archaic and of small size,
both of which characteristics commonly led, as

will

be recognized by students of Greek archi-

tecture, to a comparatively large acroterion or-

nament.

In the restoration of gutter anthemion


and central acroterion, the indications of decorative style afforded by the known antefix have
been followed as closely as possible.
It is

instructive to contrast the heavy aero-

terion of poros stone

from the temple of Assos


with the graceful and refined ornament of marble

'""'

'

'

'

'

'

'

'

'

pic. 63. Restoration of

'

''"

the
Central Acroterion, the
remaining Frag.ment dotted and shaded.

ARCH^OLOGICAL INSTITUTE.

272

regula would in that case have been cut upon the left-hand
end, remaining intact, in order to
to

the lap.

actual

On

make up

the complement

the other hand, the slight excess of the

member was

evidently determined by the desire of

the masons to diminish the length of the corner block, which

adjoined

it

upon

this side.

The corresponding lintel of the


we may safely conclude

eastern front being already known,

that the relief of the lion and the boar, one of the finest in

point of execution, was situated above the intercolumniation


of the western facade second from the north, having on
right the heraldic sphinxes, and on

its left

the missing corner

block which would have figured as the fourth in our

This conclusion
relief

is

borne out by the position

was discovered,

close

to

its

end

the western

list.

which the

in

of

the

temple, and directly opposite the intercolumniation to which


it

has been assigned.^

The mass

of rude mediaeval

masonry

in which the stone was embodied was found to consist almost

wholly of cornice blocks from the western entablature and

This circumstance would of

gable.

determine the original position of the

Among

itself

almost

the smaller fragments in the Louvre

that epistyle which

suflfice

to

relief.
is

a portion of

was situated above the westernmost

inter-

columniation of the northern side of the building, and over-

This

lapped the missing corner of the western fagade.


relief of

Upon

it

the lion springing upon the back of a hind.


is

seen, extending

from the end surface

is

the

(Fig, 65.)

at the right

across three quarters of the short length remaining, an entire


which occupied the corresponding position upon the temple of Aigina.

As

the

present writer has become aware through a study of the fragments of the latter

member, now preserved

in the

Glyptothek of Munich, none of the published

torations of the Aiginetan acroterion are altogether correct

still

res-

they suffice,

such comparison, to illustrate the characteristic differences between European and Asiatic, between advancing and provincial, design of one and the
same age.
1 Preliminary Report, plate 2. N.
in

IhXVESTIGATlOA'S
regula, fully equal to the

triglyphs to which

it

AT

ASSOS,

known width

it

is

When

is

of one of the corner

The

relief

of the southern side

evident that this lintel can only have been

situated diagonally opposite to

ner lap

273

must have corresponded.

which formed the easternmost corner


being known,

1SS3.

the only one of

its

it.

This block with the

kind which has come to

found by the French explorers

it

must have

cor-

light.

clearly

displayed the method followed by the ancient builders in cutting the lap with reference to the narrow soffit and peculiar

Fig. 65.

Epistyle Block above the westernmost Intercolumniation


OF THE Northern Side.
Series relating to the

boss of the epistyle beams.

Erymanthian Boar.

Hence

it

a matter of great

is

regret that, by the sawing of the lintel to a thin slab, these


indications have been altogether obliterated, while no record

was kept

of the original formation of a

member

so important

in architectural respects.

As

the three corner blocks of the fronts preserved to us

represent three of the deeds of Herakles,

sume

The

it is

natural to as-

was of a similar nature.


missing
adjoining the
scene, and doubtless stand-

that the subject of the fourth

reliefs

ing in connection with


wild beasts,

the

it,

lion

display the struggle of a

and boar, and the

" ces

animaux feroces dechirant

dont

la velocite

ne peut

les

lion

number

of

and hind,

les paisibles hdtes

des forets,

d^rober a leur insatiable soif de


18

ARCHAEOLOGICAL INSTITUTE.

274

Hence, the

sang," as Clarac poetically describes them.^

sumption

combat

of

as-

ready at hand, that the missing scene was the

is

Herakles with the

manthian boar.

It

Nemean

or with the Ery-

lion,

has been observed that two of the chief

same expedition,

subjects were connected with one and the

and as the adventure

of the

hero with the centaurs of Mount

Pholoe was an episode of the pursuit of the Erymanthian


boar, the latter suggestion enjoys the greater probability.

Three of the remaining


series of representations,

reliefs evidently

belong to the same

and may be supposed

have orna-

to

Epistyle Block from the Western Group of

FiG. 66.

THE Northern
Series relating to the

Side.

Erymanthian Boar.

mented the northwestern corner of the edifice. These are


the two lions and stag (Fig. 66), the lion and bull (Fig. 68),
and the lion with the legs of a deer thrown over his back
(Fig. 69).

That the

buiMing.

rendered wellnigh

is

the

small size of

48 cm.

in

length.

first

central

of these

was upon the

regula,

side of the

by the exceptionally

certain

which measures scarcely

In calculating the width of the inter-

columniation to which

it

we

appertained,

find

the distance

from the middle of the central regula to the axis of the

umn upon

the

left

surface, to be equal to 1.34 m.


to have

Thus,

if

the block be assumed

been above an inner side intercolumniation


1

col-

hand, about 7 cm. beyond the actual joint

Clarac, Musie, vol.

ii.,

seconde

partie.

of 2.44

m.

/ATEST/GAT/OA'S
on centres, as
is

is

AT ASS OS,

1SS3.

275

indicated by the dotted lines of Figure 66,

it

evident that the central regula has been displaced at least

12 cm.

Although

the right.

to

irregularities

than this have been observed in other

lintels,

even greater
this feature,

taken together with others capable of a like interpretation,


leads us to the conviction that this relief of the two lions and

stag

is

the left-hand half of that side corner panel of which

the other end

deer shown

is

in

represented by the fragment of the lion and

Figure 65.

The

half-regula at the

than 20 cm. long, being the smallest

Fig. 67.

member

left is less

of the kind

Epistyle Block from the westernmost IntercolumniaTioN OF the Northern Side.

Combination of Fragments shown in Figures 65 and 56.


to the Erymanthian Boar.

known, with the exception

Series relating

occupying the correspond-

of that

ing position upon the other side corner block of the advan-

cing centaurs, Figure 59.


to

Reference has already been made

the practical considerations which in the case of lintels

placed above side corners led to such curtailment of the inner


half-regulas.

Shorter stones could thus be utilized than would

be possible had the joint surfaces conformed to the axes of the


columns.

The

small size of the detail in question

may hence

serve as an argument in support of the combination of the

two fragments shown

in

Figure 67.

Computing the length

of

the panel thus constituted, and assuming the two metopes

above

it

to

have been of equal

size, w^e find

joint surface to corner of lap, to

the stone, from

have been of a dimension

ARCHAEOLOGICAL INSTITUTE.

276

closely agreeing with the 2.9

pying

this

manner

for the lintel occu-

In combining the fragments

position.

indicated,

m. required

we

the

in

are struck with the agreement in de-

sign of the groups thus brought together.

Just sufficient

space remains upon the missing portion to contain the trunk


and hind quarters of the lion whose head and shoulders are

The

seen upon Figure 65.

beasts of prey spring

upon

their

booty in the same way, and with bodies inclined in the same
direction, forming in their conventional regularity a pendant

Fig. 68.

Epistyle Block from the Western Group


OF THE Northern Side.
Series relating to the

to the

monotonous

other end

file

Erymanthian Boar.

of the advancing centaurs

upon the

of the building.

The second

relief

belonging to this series represents a lion

dragging a bull to the ground.

(Fig. 6%?)

That

this is to

be assigned to one of the inner intercolumniations of the


sides

is

susceptible of

little

doubt, as the distance between

the two remaining regulas exactly corresponds to the dimen-

had the two metopes been


and the triglyphs of the normal side width.

sions requisite for such a lintel

equal in

The

si2e,

half-regula at the right

quite possible that the block

is

slightly excessive,

may have

ner panel, shown in Figure 6^.

and

it

is

adjoined the side cor-

In this case the third

tri-

glyph from the northeastern corner, though slightly broader


than the second, and than that above one of the epistyles of

AT

INVESTIGATIONS
the southern entablature, would

ASSOS,

1883.

277

have been two or three

still

centimeters narrower than the average.


In regard to the third

relief,

lion with the legs of a deer

we have

that of the hind quarters of a

thrown over

a further indication of

its

his

back (Fig. 69),

connection with the series

of the northwestern corner in the position in

which

it

was

found, immediately adjoining the relief of the lion and boar,^

The

only measurement which can be taken from the small

fragment
excess

is

that of the half-regula, 33 cm. in length.

is

here so great that

we

The

are led
33

to consider the possibility of this block

being the end of the corner

^"

r'J^rj~~

the

lintel of

western facade, complementary to the


Its

lap.

evident superiority in

design

and execution might be put forward


support of this identification.
the other hand,
,r.

we cannot

-1

<-

the entire smiilarity of


scale

fail to

1.1

in

But, on

J^'S

observe

Fig. 69.

Fragment of

AN Epistyle Block of
^^ series relating
'^ ^he Erymanthian

^JZ3~[~''"'i

the anmials in

and action with those sculptured

Boar.

upon the two blocks of the northern


side, and the want of any indication of a struggle of either
boar or lion with Herakles.

The

stone represents too small

a portion of the panel to allow of any decision in this matter;

and
to

it

might, upon grounds quite as convincing, be deemed

have formed the

left half

of the side block of the lion and

bull (Fig. 68).

There remain but two sculptured epistyle blocks, which


cannot be brought into connection with either of the four
chief scenes depicted upon the corners of the temple.

These

are the reliefs of the butting bulls, Figures 70 and 71.


Bearing in mind the perfectly regular manner in which the se-

quence

of subject
^

was carried out

in all the reliefs of the

Pirliminarv Report, plate

2,

N.

ARCHAEOLOGICAL INSTITUTE.

78

entablature

considered,

hitherto

sequence in favor of

which the fundamental law of the Doric

style forbidding the

decoration of the epistyle had been sacrificed,

it

appeared

at first sight impossible to assign a position to these blocks.

On

the one hand,

was

it

difficult

to conceive

their

having

adjoined either of the series depicting the deeds of fierakles

on the other, the number

lateral

of

epistyle

beams being

even, there was no possibility of the bulls having occupied


the middle of the sides, in that 'absolute

symmetry and

dis-

connection from the other scenes which would have been

Fig. 70.

Epistyle Block from the Entablature of the Cella,


ABOVE the PrONAOS.
Subject related to the Cult of Assian Athena.

demanded by the subject and composition of these reliefs.


Such difficulties are, however, entirely avoided by their removal from the peripteros. That these two epistyle blocks
were, in fact, situated within the colonnade, upon the wall
of the cella, can

be proved from indications of so different

nature that the argument

is

vicious circle against which

guard

free

from those dangers of a

we must

constantly be on our

in reconstructions of this kind.

One

of the reliefs,

being the only

shown

lintel,

which the extent of

all

Figure
its

70,

is

remarkable as

entire length,

the mouldings can be measured.

half-regulas, respectively 31
cessive.

in

preserved in

upon
Its

and 37 cm. long, are both ex-

Deducting from the actual length

of the stone the

INVESTIGATIONS AT ASSOS,
we

i6 cm. thus determined,

1S8S.

2/9

find the axes of the

columnar

triglyphs to have been only 2.35 m. apart, or 9 cm. less than


the average of the side intercolumniations.
In like manner,

the fragmentary relief of the

but

same subject

(Fig. 71)

measures

14 m. from the remaining joint surface to the middle of


the central regula.
The span thus indicated is not less than
1.

16 cm. narrower than that of the side lintels; and although


this dimension was without doubt slightly increased through
a want of correspondence between joint surface and ideal

Fig. 71.

Epistyle Block forming the Pendant to that

SHOWN
Same
axis,

IN

Figure

70.

subject and corresponding location.

we have good grounds

for believing that the

amount

of

such correction cannot have exceeded six or eight centimeters


in the total length.

This

lintel

is

unfortunately deprived of

both half-regulas, so that no direct information

is

upon

this

the metopes

above

it

point from the stone itself

be assumed

the joint surface

is

to

But

if

have equalled those above

found

to

obtainable

its

pendant,

have been removed from the axis

of the support by not

more than three or four centimeters.


same conclusion is furnished by
the subjects represented upon these panels.
The animals are

An

indication leading to the

so entirely similar in posture and proportions that the outlines actually appear to have been transferred to the stones
from one and the same drawing.
This may be tested by

laying a strip of tracing-paper. bearing the outlines of P'igure

ARCH.'EOLOGICAL INSTITUTE.

28o

yo over Figure
a

in

71,

when

manner otherwise

the forms

will

inexplicable.

be seen to agree

The only

variations

are those rendered necessary by the different height of the


tainias, this
half,

member measunng

which alone we
Hence,

tical.

in

one case but eight and a

the other fully ten centimeters.

in

The

lengths, with

are at present concerned, are absolutely iden-

it is

impossible to assume that the figures upon

the fragmentary block occupied a more elongated panel.


It is

thus evident that

we have

to deal with lintels placed

above considerably narrower intercolumniations than those


of the sides, which, with the exception of the corners,^ are

known, from the traces


have

columns upon the

stylobate, to

no case varied more than a centimeter or two from

in

the average of 2.44 m.


bulls

of the

In short, the reliefs of the butting

formed no part whatever of the entablature

They can only be assigned

peripteros.

to the

of

two side

columniations of the pronaos, between the antae and the

umns

col-

lintels

above these openings

only ones of a smaller span

than those hitherto

in antis,

are the

the

inter-

inasmuch as the

considered.

The

exact widths of these intercolumniations are not ascer-

upon the stylobate being


permit of an accurate measurement. The

tainable from the plan, the traces

too indistinct to

columns

in antis stood

upon blocks

larger,

and without doubt

deeper, than those of the surrounding pavement, having ex-

posed surfaces of
^

The

1.2

m. square and

i.i

by

1.4 m.

(Compare

deviations of the corner intercolumniations of the sides from the nor-

mal width, evident from traces upon

tlie

stylobate referred to

upon page 76

of

the present volume, are themselves not equal to the difference in length between

the epistyle blocks sculptured with the two bulls and the lintels placed above
the inner intercolumniations of the sides.
these traces

is

The

increased span indicated by

readily explicable by the consideration that the corner passages

while it is,
of the sides were intentionally approximated to those of the fronts
on the other hand, impossible to assume that a restriction of such amount can
;

have existed

in the case of

any intercoluniniation of the peripteros.

INVESTIGATIOXS
the plan, Fig.

Upon

4.)

AT ASSOS,

28 I

1SS3.

these blocks are to be seen the dis-

colored patches in slight relief resulting from the presence

upon them

for

wellnigh two thousand years of the lowest drums

The

of the shafts.

outlines of the arrises, here eighteen in

number, could be followed

in but

one

and even there

case,

with no certainty in respect to the demarcation.

All that

could be definitely determined was that the circular patches

occupied about the middle of the slabs, in their axes from

The

east to west.
to

have been

central opening

at least

1.9

may

in this

and possibly 2 m.

wise be seen

in the clear,

the columns in antis 2.8 or 2.9 m. on centres.

and

This shows

the side intercolumniations to have been spanned, from centre


of anta wall to axis of column, by lintels but 2.2 or 2.25 m. in
length,

and

to

have been even narrower than we should have

been led to assume from the dimensions of the blocks now


under consideration.

It

is,

however, probable that the ad-

justment of the members of the

ment

frieze led to

some

displace-

of the columnar triglyphs from the axes of the supports

Although the central intercolumniation was

beneath them.
thirty, or

even

forty,

centimeters wider than the intercolumni-

ations of the fronts, the

average of the three pronaos

was almost exactly equal

to that of the

lintels

side lintels of the

Thus the average width of the metopes and triby the dimensions of the outer entablature, would naturally have been retained, and the axes of
these members have tended to a displacement in the sense

peripteros.

glyphs, determined

indicated.

This

is

precisely the conclusion which might have

been drawn from the exceptionally short distance between


the columnar axes and the excessive lengths of the halfregulas

upon these blocks of the butting

bulls.

Lintels of

the normal length supplied by the quarrymen were, from


constructive reasons presently to be adduced, adjusted with-

out curtailment to agree with

the details of a frieze cor-

ARCH^OLOGICAL IXSTITUTE.

282

responding to somewhat narrower intercolumniations than


those of the peripteros.

any

obviously impossible to assume

It is

fundamentally different division

of the

pronaos

Broad as the central intercolumniation was, the


it

was

much

still

lintel

frieze.

above

too short to permit of the introduction of

metope and

a supernumerary

We

triglyph.

need not, of

course, suppose the spacing to have been perfectly equal

the increase of ten or fifteen centimeters, apparent from the

would have

lintels in question,

tion

in

manner

would tolerate the


the fagades.

sufficed to eff'ect the equaliza-

satisfactory

perfectly

irregularities

known

any eye which

to

to

have existed upon

In the light thus thrown upon the construction

of the cella front,

we can

readily perceive the reasons

which

led the ancient builders to permit a very considerable exten-

sion of these lateral epistyle

columns

in antis.

beams beyond the axes

of the

This can only have been done for the pur-

pose of decreasing the length of the central

lintel

spanning

the widest intercolumniation of the building.


It

may

further be noticed that the two epistyle

beams with

the reliefs of the butting bulls are the only ones which do not
display upon the edges of their soffits those shallow sinkings
of rectangular plan intended to receive the

during the process of shifting the stone to

Although scarcely

end of a crowbar

its

exact position.

of itself sufficient to furnish a definite proof,

this peculiarity suggests the

employment

of

another means for

the lifting and setting of these lintels of the pronaos than that

adopted to meet the different requirements of the peripteros,

and may hence be advanced

in

support of the arrangement

proposed upon other grounds.

That the entablature upon the front of the


have been distinguished by sculptures,

is

cella

should

entirely in keeping

with the importance assigned by ancient architects to the


portal of the

inner fane.

Doric temples, in which sculp-

AT

IXVESTIGATIOXS

ASSOS,

2S o

1S83.

tured decorations were, from motives of economy, not extend-

ed to the entire

edifice, invariably display their finest reliefs,

whether metopes or consecutive

Well known

instances

this

of

frieze,

are

the

above the

pr.onaos.

temple of Bassai,

the temple of Sounion, and that most striking analogy, the

Theseion.

Moreover, the subject represented upon these two lintels,


so entirely unconnected with the extended scenes of the
outer entablature,

is

such as

to

render them eminently suit-

able to the decoration of the inner house of the goddess.

It

relates, not to the exploits of the Aeolic hero there depicted,

but, as will be set forth in a subsequent connection, to the

peculiar cult of Trojan and Assiau Athena.

And,

in conclusion, the

reference

manner

to

in

the

strictly

argument may be enforced by a


symmetrical and almost heraldic

which the animals are depicted.

It

is

at

once

obvious that these panels were intended to be seen as pendants.

So decisive

the indication thus afforded, that

is

it

induced the designer of the French restoration of the temple


to assign these shortest known epistyle beams to the widest
intercolumniations of the building, at the corners of the main
fagade.i

j^ jg

impossible not to recognize the correctness of

the instinct which led to this error, readily to be detected by

measurement of even the few fragments removed to the


Louvre and it is satisfactory that the principle, though in a

different application,

may now

be justified by the most careful

comparison of the actual dimensions.


1

Texier, Description, vol.

ii.

pi.

12.

In the volume of VUnivers, referred

to in a foregoing note, the author explicitly remarks, "


la

Les deux extremites de

fa9ade etaient ornees de deux couples de taureaux dans I'attitude de combat."

The

lion

and

In order to

bull,

and the two

make them

fit

lions

and deer, occupy the intermediate

lintels.

the given spaces, these blocks have been subjected to

a truly Procrustean elongation, even the reliefs of the wild beasts last mentioned being of that short length which proves them to have belonged to the
sides of the building.

ARCH^OLOGICAL INSTITUTE.

284

now

Six sculptured metopes belonging to the temple are

known, three of these having been removed from Assos to


the Louvre, and three brought to light during the American

Four are preserved

excavations.

in their entire width, yet in

only one of these can the exact distance between the edges

This

of the bordering triglyphs be ascertained.

is

due

to the

fact that rebates, of slightly variable depth, were generally

cut

upon the ends

of the projecting bands, so as to

fit

into

the corresponding reveal of the triglyphs, in order to prevent


In such cases the exposed

the appearance of open joints.

face was, of course, narrower than

the slab

itself,

and no

conclusions are to be drawn in this regard concerning four


of the metopes, the edges of

which have been so weathered

and defaced that the extent of the overlap cannot now be


determined.

Accurate measurements
from the
total

relief of the

sion

is

be taken only

While the

running centaur. Figure 72.

width of the slab

face visible

in this respect are to

is

a fraction over

between the rebates

is

']},

cm., that of the

The

68 cm.

dimen-

latter

exactly that of the exceptionally narrow space remain-

ing between the regulas of one of the epistyle blocks of the


eastern front

and there can be

little

doubt that

was placed above the right-hand side

of

metope

this

the second inter-

columniation, from the south, of this facade, as

shown

in

Figures 55 and 62. The sculptures upon the epistyle, and


the metope above it, are thus seen to have agreed, in this
instance, both in subject
is

and direction

natural to assume that this

of

composition

was the case with

metopes where such an agreement

is

all

and

it

those

indicated by repetition

of type.

According to

this principle, the relief of the

two squatting

sphinxes repeated in heraldic symmetry. Figure

']'i^,

would be

assigned to the central intercolumniation of either the eastern

INVESTIGATIONS

AT ASSOS,

2S5

1883.

or western front.

Unfortunately, the width of but one of the


metopes placed above the epistyle blocks in question can
be measured from the lintels, while the edges of the sphinx
metope are so rounded that it is impossible to determine the
exact width of the exposed face of this slab.
of proving this contiguity are at hand.

Hence, no means

The dimensions

as

they can be ascertained do not, however, render it


improbable. The width of the left-hand metope, above the
central epistyle of the eastern front, was
75 cm., somewhat
far as

Fig

73-

Fig. 74.

Metope of the Eastern Entablature occupying the


Fourth Field from the South.
Related

in

subject to the series of the Centaurs of

Mount Pholoe.

Fig. Ti.
Metope showing the Coat of Arms of Assos.
Probably placed above the central epistyle block of one of the
fagades.

Fig. 74.

Metope related in subject to the


Erymanthtan Boar.

series

of the

than that indicated by the Z2 cm. of the sphinx metope,


if the rebate upon this slab be assumed
to have been of the
same depth as those upon the two centaur metopes.
The
less

right-hand metope above the central intercolumniation of


the
eastern front was undoubtedly somewhat wider than the
left,
averaging, together with the two metopes following
towards
the north, above the missing epistyle block, fully
"jG cm.
In like manner the metope of the boar, Figure
74, agreeing
as it does with one of the reliefs of the
western front, both
in subject

and direction of composition, may be supposed to

ARCH^OLOGICAL IXSTITUTE.

286

have occupied one of the


north,

of the frieze of

first

four panels, counting from the

that facade.

Here, again, proof

is

lacking, as the edges of the slab are rounded, and the space
of the relief of the lion

between two of the regulas


cannot be measured.

We

and boar

must, furthermore, admit the pos-

metope having been above the missing lintel


corner, the subject of which was in all
northwestern
of the
likelihood that of Herakles and the Erymanthian boar.
sibility of this

In the case of the three metopes brought to light by the


American excavations, we have such additional information as

Fig

Fig- 75-

Fig. 75.

76.

Fragment of a Metope, related in subject to the Series of


THE Centaurs of Mount Pholoe.
Probably from the southern half of the eastern entablature.

Fig. 76.

Metope of uncertain location and

Fig. 77.

Fragmentary Metope, of uncertain location and

may be

derived from a record of the position in which these

reliefs

were found.

The

subject.

small fragment showing the hind

legs of a centaur, Figure 75,

was unearthed close

eastern corner of the temple, together with a

known
There

to
is

subject.

to the south-

number

of blocks

have appertained to that part of the entablature.


thus good ground for the belief that this centaur

metope, like that other one already considered, stood in connection with the series relating to Herakles and Pholos, and

occupied one of the three southernmost panels of the eastern


facade.

It will

be observed that the galloping position and

INVESTIGATIONS AT ASSOS,

287

1SS3.

the direction of the course are in both cases the same as


those of the centaurs depicted upon the epistyle.

The entire metope, Figure y^, was found to have been removed from the spot where it had fallen, and to have been
incorporated into the

rude Byzantine fortifications erected

upon the west

>of the citadel.

plan

upon

Plate

the

marked
Report,

reliefs

found

this

in

have belonged to the western facade,

to

appears probable that this metope ornamented that end

was not placed above the relief


Herakles and Triton, the regulas of which correspond to

of the building.
of

It certainly

slabs of considerably greater width

the composition, from


it

is

Preliminary

exact point

Inasmuch as both the other

2.

rampart are known


it

The

Acropolis,

the

of

left

was above the southern half

The two fragments


found buried

in

is

and the movement of


it

unlikely that

of the eastern front.

metope shown

of a

in

Figure

J"]

were

the earth which had accumulated upon the

stylobate of the temple

the slab

to right, renders

itself,

not apparent,

we

and, as the original width of

possess no indications whatever

in regard to its original position.

The two metopes

last

considered apparently represent com-

bats of Herakles, and form further illustrations of the theme

chosen

for the decoration of the peripteros.

Standing

in

no

direct connection with the four great scenes of the epistyle,

they open a wide

field for

conjecture in regard to the

of exploits thus depicted.

absolutely impossible, that

It
all

is

number

highly improbable,

if

six of these reliefs are to

not

be

assigned to the metopes of the eastern front, the widths of

which, enumerating them in their order from south to north,

known

have been 93, 74, 81, 68, 75, three unknown


averaging ^6, and, finally, 90 and 85 cm. It is hence to be
assumed that all the metopes of both fronts were sculptured,
are

to

and that fourteen of the twenty slabs thus distinguished have

ARCHyEOLOGICAL INSTITUTE.

288

That a proportionately smaller number

disappeared.

of

meto-

pes than of epistyle blocks should have survived the ruin


of the edifice,

is

readily explained

their

them better

by the consideration that

shape, and plane

smaller size, square

backs adapted

needs of Byzantine and Turkish builders.

to the

Restriction of the sculptured metopes to one or both fronts

was a common practice

Greek

architecture, as, for in-

The metopes of
may be supposed to have

stance, in two of the temples of Selinous.

the sides, like those of the fronts,

been painted, the differences between the

reliefs

and the

unsculptured surfaces being rendered less apparent by this

means.

Here we may terminate our investigation


occupied by the sculptures of the temple.
those

who may

into the positions

For the benefit

not have cared to enter into

all

of

the details of

the argument, the conclusions are graphically set forth in the

plan of the entablature, Figure

now known

blocks
of

their

subjects

shown

are

are

given

The

'j'i.

sculptured epistyle

in solid black,
in

the

larger

and the names

lettering.

The

smaller lettering indicates the probable sequence of the series.

The arrows drawn between

the lettering and the entablature

denote the direction of the composition of the sculptured


blocks, and

show the

entire regularity of the

arrangement

in

this respect.

The

question which naturally arises as to the proportion of

sculptured blocks,

known and unknown, is best answered by


One lintel is certainly lacking from

an analysis of this plan.

the eastern, and two from the western facade.

If the reliefs

were extended over an equal number of intercolumniations

upon the two

sides, at least six sculptured epistyle blocks of

the shorter span are missing.


plore

is

The

loss

which we most de-

that of the central lintel of the pronaos entablature,

the subject of which doubtless had reference to the cult of

oo
LU

X
^
T

H
DC

CO

^
Q

Qi

D
<
H
Z
LU
u

(/^

O
_i
o
X

Q-

o
z

C/)

_1

<

LU
LU

<
LU

HESIONE MYTH

ADVANCING CENTAURS

HESIONE MYTH

ADVANCING CENTAURS

HESIONE MYTH

ADVANCING CENTAURS

WILD

beasts]

HIPPOLYTE MYTH

WILD

BEASTS

HIPPOLYTE MYTH

<
5c

z
<

lu
1

Fig. 78.

UU

CO
LU

X
CO

<

UJ
CQ

X
a.

c/^

uu

o
Q.

<

a:

n.

Q
O
<J

CO
uu
1

<
CC

Plan of the Epistyle of the Temple of Assos, showing


IN Black the Position of the known Reliefs.

ARCH^OLOGICAL INSTITUTE.

290

From

Assian Athena.

assume the epistyle

these imperfect indications


peripteros

the

of

sculptured panels, of which twelve

Of

considerable fragments.

to

now

we should

have had twenty

exist, in

more

or less

the unsculptured epistyle blocks,

the remains of eleven, or possibly of twelve, were found upon


the

If

site.

twenty-four of the forty-four

teros were plain, this

that of the

known

lintels of the perip-

a somewhat smaller proportion than

is

to the

unknown

reliefs.

It is,

however, to

be supposed that the stones with plane faces would be

and Turkish builders

lected by Byzantine
others,

the

Mohammedan

se-

in preference to the

antipathy to graven images de-

termining this preference no less than the practical considerations of unskilled masons.

The most
this

significant fact

examination

is,

was not broken up


as has been

which has been

elicited

through

that the sculptured face of the epistyle

independent and disconnected panels,

into

assumed

in the

French

restoration,

and

in all

those text-books which treat of this subject, so exceptional in

The

the history of Greek architecture.

reliefs

extended compositions, with their chief scenes

formed four

at the corners

of the building, separated in the middle of the facades by the


coat-of-arms of the city.

canon

of the

Doric

style,

embody-

ing one of the principles of architectural propriety, forbade


the decoration of

members

directly functional in the construct-

ive framework, but, in this instance, this

canon was

set aside

for a definite purpose, that of securing for the sculptures the

continuity of the Ionic zophoros.

It

was a

similar purpose,

coupled with the determination to respect the laws of the


style,

which induced Iktinos

to

hide the frieze of the Par-

thenon behind the columns of his peripteros.


respects

we must condemn

temple of Assos.

much maligned

the expedient

Yet, in tardy justice to the

architect,

we may now

In aesthetic

adopted in the

unknown and

recognize the fact that

lAVESTIGAT/OXS AT ASSOS,

1883.

29

the decoration of the epistyle was decided upon, not as an


altogether purposeless innovation, but in the desire of secur-

ing a recognized advantage

and, as in every genuine en-

deavor to effect an improvement, the design, with all its


faults, thereby attained a high degree of individuality and
interest.
In connection with a plan of absolute regularity,
this provincial

decorations,

have given

attempt to extend and connect the sculptured

undisciplined

and injudicious as

it

was, must

the temple

an air of rude picturesqueness


analogous to that which we find so attractive in the earliest
to

edifices of the

Romanic

style in

Northern Europe.

CHAPTER

IV.

DATE OF THE TEMPLE.

THE Temple

of

Assos

is

of unique interest in the history

because of the enigmatic char-

of ancient art, not only

acter of

only

its

sculptures, but because of the fact that

known Doric

decadence

of

edifice in

of

its

date

therefore a matter of signal importance.

ment

this regard

of the disputes

the

Asia Minor which antedates the

The determination

that style.

it is

which have arisen

in

definite settle-

would

American expedition

of itself justify the direction of the

is

to

this site.

Prior to these investigations


tive nature

was

little

information of a posi-

to be gathered, either from the fragmentary

and disconnected

reliefs

removed

from the

to Paris, or

gether incorrect and misleading engravings


tectural

remains

explorers.^

ment and

which

had

Notwithstanding
its

position

in

been
this,

by

published

age of

the

development

the

the

of

earlier

the
of

alto-

archi-

monu-

Hellenic

sculpture and architecture have been discussed with extra-

ordinary boldness of assumption by those scholars whose


delight

it

is

to

reduce every phenomenon

world to a categorized system.

who

Almost every

classic

archaeologist

has had occasion to refer to the Assos reliefs stands

committed

to

some opinion

expressed with greater or


^

the

of

on

less

these

points.

precision,

is

to

date,

be found

Instances of Texier's perversions of fact in his description of the architec-

ture of the temple have been adduced in sufficient

Report, pp. i8, 99, etc.

number

in the

Preliminary

INVESTIGATIOXS AT ASSOS,

1SS3.

293

almost every book which deals with the history of ancient

in

was

It

art.

inevitable that such assumptions should

by a high degree of incompatibility, and we have

acterized

cause for surprise when

tle

be char-

the twelfth century to the

we

century before Christ,

fifth

lit-

them ranging even from

find

semi-mythical ages preceding the Dorian migration

from
to

the

years which witnessed the rivalry of full-grown Hellenic states,

and the outbreak


Bursian

of the Peloponnesian

Mykenai and with works

of the Lions at

the period of
history, Perry
acter,

be "

War.

brackets the sculptures of Assos with the Gate

its
2

of

Assyrian art at

highest development.

In a more recent

their "

most primitive char-

upon

lays stress

... in the highest degree archaic," conceiving them to


among the very earliest works of the relief style."

Friedrichs,^ followed

by Wolters,^

refers

works of Greek

them

among

century, remarking that they are

to the seventh

the most ancient

and that they cannot have been preceded by any development extending over a considerable
period

of

art,

Studniczka^ assigns

time.

seventh century

first

later
^

them with monuments

half of the sixth century

s.

v. Griechische

to
"

the

hoch-

of the seventh

Murray' thinks them not

than the middle of the sixth century

Bursian,

reliefs

Furtwangler,^ describing them as

alterthijmlich," classes

and

the

and Liibke^ char-

Kunst, in Ersch und Gruber's AUgemeine Encyclo-

pddie, vol. Ixxxii., Leipzig, 1862.


2

Perry (Walter Copland), Greek and

Friedrichs

1868-71, vol.

(Carl),
i.,

and

in

Diisseldorf, 186^-79, vol.

Berlht's

Roman

Antike

Sculpttire,

Bildwerke

London,

{Bausteine),

1882.

Diisseldorf,

Schnaase's Geschichte der Bildenden Kiinste, 2d


ii.

ed.,

p. 126.

Wolters,

in Friedrichs's Gypsabgusse antiker Bildwerke in historischer Folge


2d ed., Berlin, 1885.
5 Studniczka, Attische Porosgiebel, Mittheilungen des deutschen archiologisckeft
Instituts, vol. xi., Athen, 1886, p. 75.
*

erkldrt,

Furtwangler,

Murray, History of Greek Sculpture, London, 188J-83.


Lubke, Geschichte der Plastik, 3d ed., Leipzig, 1S80, vol.

s.

v.

Herakles, in Roscher's Lexikon, p. 2193.

i.

ARCH^OLOGICAL INSTITUTE.

2 94

acterizes

them

sculpture which have


Krell

refers

most primitive among those works

as the

come down

building

the

from the archaic age.

to us

the temple

of

of

the

to

period

immediately succeeding the Lydian conquest of the Troad


(b.

560), but adds that, judging from the sculptures,

c.

may be somewhat more

recent

it

while Durm,^ expressing a

decided opinion that the building was anterior to the invasion


of Kroisos, fixes the date, with a single note of interrogation,

year 600

at the

made

b. c.

in the case of the

certain allowance

two authorities

as these viewed the problem mainly in


pects.
their

The drawings and

is,

however, to be

named, inasmuch

last

architectural as-

its

restorations of Texier, upon which

judgments were based, undoubtedly do bestow an

extreme rudeness upon the temple.


Krell terms
reality

it,

"

perfectly

baggy

" entasis of

The

the shafts (which are in

straight-lined), the

mouldings interposed between the

air of

excessive, and, as

fictitious

frieze

course

and the corona,

of
in

strange disaccord with the normal forms of the Doric entablature,

the

supernumerary steps upon the

fronts,

the lack of an incision between hypotrachelion and upper-

most drum,

the trapeze-shaped

regulas,

combine with

host of similarly incorrect features of plan and elevation to


give to the edifice a grotesque and primitive aspect, which

would, indeed, be wholly inexplicable in any


of

Greek architectural

known

period

history.

Contrasted with these extreme views,

we have

the opinions

of two historians of Greek art, whose writings have been


distinguished in a high degree by penetration, lucidity, and

independence, and

who have been

the

first

to

advance well

Krell (Philipp R), Geschichte des doriscken Styls, Stuttgart, 1S70, p. 20.

"

Durm, Bankunst dcr

"pronounced

Obvi5, and p. 135.


word which he employs, Durm speaks of the

Gricchen, Darmstadt, 1881, p.

ously mistaking the sense of the

archaistic \sic\ character " of the temple of Assos.

LWESTIGATIONS AT ASSOS,

18S3.

295

founded doubts concerning the great antiquity so generally

The

attributed to the sculptures of Assos.

Reber,^ has contended,

in

those of Furtvvangler, that


thiimlicher Fries " does

earlier of these,

words exactly contradictory


the " keinesvvegs

so

to

hochalter-

not warrant the assignment of the

temple of Assos to the most archaic period of the Doric


style.

this

The

second, Overbeck,^ quoting Reber, has enforced

view with definite arguments, carrying the examination

He

into greater detail.

analogous to the

has pointed out that representations

relief of

Herakles and Triton,

style as well as of subject, are found, not


of

immature Greek

point of

in

among

the works

but in a class of vase paintings of

art,

comparatively late date, remarkable because of the frequency

with which they depict figures with inclined bodies.


class of vase

paintings he maintains to be the earliest in

which such ingeniously designed subjects,


in

organic

ment,

are

respects, albeit

known

to

appear.^

smaller figures of the same

horse-legged

centaurs

fully

somewhat exaggerated
Indications

parallels such as this, from the formation

of

This

relief,

upon

developed
in

movefrom

derived

and action

of the

and from the appearance


another

block,

have

led

Reber (Franz), Geschichte der Baukunst im AltertJmme, Leipzig, 1866,


Kiinstgi'schichte des Alterthums, Leipzig, 1871, p. 213.
In the American
edition of the latter work, History of A)icient Art, by Dr. Franz von Reber,
translated and augmented by Joseph Thacher Clarke, New York, 1882, the
translator omitted the words above quoted, thinking it advisable to be less
1

and

committal in respect to the age of the sculptures, in view of the decisive information so soon to be expected from the excavations at Assos, then about to
.commence. This omission he has now cause to regret.
2

Overbeck (Johannes), Geschichte der griechischen

Plastik,

3d

ed., Leipzig,

18S1-83.
8 The chief argument to this effect is based upon the fact that designs of this
nature are not to be found in archaic compositions of the style represented, for
instance, on the vases published in plates 95 and 96 of Gerhard's Auserlesene

Grieckische Vasenbilder, while they correspond well with later paintings, such
as those
plate

shown on

plates 94, 102, 105,

of Gerhard's Etruskische

in, and 113

und Kampaniscke

of the

same work, and on

Vasenbilder.

ARCHAEOLOGICAL hXSTITUTE.

296
Overbeck

to

deny

doubt whether they


tieth

the Assos sculptures that great anthem by so many antiquaries, and to

to

tiquity attributed to

are, at all events, earlier

This

Olympiad.

is

the

first

than the six-

and only attempt

to deter-

mine the age of the temple by the comparative methods


historical science

for

this respect to Clarac,^

of

we can scarcely give precedence in


who supported his opinion that the

of Assos were contemporary with the gable group of

reliefs

Aigina by arguing

informs

that, as Pliny ^

us,

it

was not

until

the fiftieth Olympiad that the Daidalian sculptors Dipoinos

and Skyllis employed marble


quently assume

that

in statuary.

We

may

conse-

hard and coarse stone of Assos

the

could not have been worked by the Greeks before the ac-

quirement of considerable experience in the tooling of more


tractable materials.

In view, however, of the scanty and

cited,

no surprise can be

actual truth.
to

at

their failure to hit

upon the

The American

investigations have

been the

provide materials fully adequate for a solution

the problem.

numerous

The

known

reliefs

and

as before,

are,

Above

now

nearly twice as

mythological series, the sub-

which were carefully considered and highly


all,

acteristics

position of
1

are

of the edifice enables


its

us to establish

the exact

design in the history of the Doric style.

Clarac, Musie, vol.

ii.

2d partie, Paris, 1841.

Clarac, as will be

of the present volume.

It

argument such as

Pliny, Nat. Hist.,

sawed

remem-

into thin

Compare note i, page 51,


seems unnecessary to enter into a serious refutation

slabs under his personal supervision, to be a granite.

significant.

a definite determination of the architectural char-

bered, believed the trachyte of the Assos reliefs, which were

of an

of

moreover, seen to have stood

in definite relation, as parts of a

jects of

the want of agreement be-

felt at

tween their estimates, or

first

whose opinions have been

disposal of the scholars

at the

untrustworthy data

this.

XXXVI.

4. i.

AT ASS OS,

INVESTIGATIO.\S

1883.

297

The conclusions thus reached may be concisely stated at


The temple of Assos was erected during the age

the outset.

which had seen the termination of the Persian wars, towards


the middle of the

century before Christ, at that period

fifth

the Greeks of the Asiastic coast were in the

when

joyment

of their relief

ing was unquestionably

somewhat

en-

first

The

from Oriental oppression.

build-

later in date than either the

temple of Aigina or the Theseion, and contemporaneous with,

somewhat older

or

many advanced

appearance of the sculptures, to which


tures present a

The archaic

than, the temple of Sounion.

marked

contrast,

is

to

fea-

be attributed to the

influence of local and conservative tendencies, favored by the


refractory character of the material in which they were executed.

Paradoxical as

of nearly the

same date

it

may

appear, the temple of Assos

is

as the Parthenon.^

That rude and primitive character, which so many writers


on Greek art have sought to explain by the assumption of a
remote date,
which, be

is

to

in the estimates of

Are we

be attributed to provincialism

observed, has by no

it

many

actually to believe,

a factor

means received due weight

similar archaeological problems.


it

may be

asked, that the rude figures of the

temple of Assos were sculptured at the same period as the incomparable gable
groups of the Parthenon } Not necessarily. We have to take into consideration in this connection the revolutionary hypothesis of Puchstein,

first

pub-

lished in the Berliner philologische IVochetisckrift, Winckelmannsfest der arc/idologischen Gesellschaft zu Berlin, January 18, 1890,

and subsequently elaborated,

with highly interesting sections of the draperies of the different schools,


fahrbicch des prenssischen archdologischen Instituts, vol.
theiionskulpturen.

According to

v.,

1890, Heft

be ascribed to some date subsequent, at

all

events, to 430 b. c,

half a century than the architectural design.

They

are to

and are possibly

If

we

are to seek,

the contemporary works of European Greece, a parallel to the sculptures

more fitting example, itself displaying marked inequality


would be the central group of the eastern gable of the temple of
Olympia, which must be very nearly of the same age.

of the temple of Assos, a


of treatment,

Zeus

the

which in the opinion of the present


methods of comparison, the sculptures

of the Parthenon were not the work of Pheidias, or of his generation.

among

in

Die Par-

this view,

writer has been established by scientific

more recent by

2,

at

ARCH^OLOGICAL INSTITUTE.

298

the point of view of the practical investigator,

From

it

is

pecuHarly unfortunate that the foundations of the temple of

Assos should have been

upon the native rock,

laid directly

and have thus precluded the

possibility of a discovery, be-

neath the stylobate, of vestiges of an earlier occupation of


the

such as those which proved of decisive importance

site,

The

the Olympian Altis.

in the exploration of

which any considerable bed

of

earth

existed

only spot in

beneath

the

stones of the temple was within the cella, and here the laying
of the mosaic pavement, apparently

to

be referred to the

fourth century before Christ, led to disturbances of the soil

which make

it

impossible to draw conclusions from objects

discovered therein respecting the epoch of the original construction.^

The

inconsiderable deposits of earth which exist

beneath the paving stones of the pteroma, and which may,

may not, have been disturbed at a subsequent period,


were, when accessible, carefully sifted by the explorers, but

or

were found

to contain

nothing beyond a few small sherds of

an unglazed pottery, such as was


earliest to the latest ages of

in

Greek

every-day use from the


antiquity.

In the entire absence of direct indications, the only definite proofs as to

the age of the building are to be derived

from a comparison of the leading features of the temple of

Assos with those

of

temples of the same style and of

Architecture, as has been generally recognized,

date.

known
is less

influenced by personal and local conditions than sculpture,

and

is

hence that
This

traced.

of building

The

is

in

art in

which a regular growth

due to the

fact that

is

best to be

development

in the art

respect to design, as well as in respect to

discovery, beneath the mosaic floor of the naos, of various fragments

and moulded vessels, as well as of a bronze coin,


all of which
probably belonged to the fourth century before Christ,
has been mentioned in
a former chapter, page 72.

of painted

AT

INVESTIGATIONS

ASSOS,

1SS3.

299

those improvements of plan and construction which are inseparably connected therewith

is

dependent upon the regular

advance of technical invention, and upon the degree

and culture

ization

of a race or

of civil-

community, rather than upon

any stimulation or refinement of the imaginative and perceptive faculties of individual

The works

nity.

members

of the formative arts, in the historical view,

are always affected in a

marked degree by

Throughout the long

influences.

commu-

of such race or

local

and individual

of artistic

vista

history,

sculpture and painting are naturally divided into countless


schools,

working

architecture
like the

ing,

on the

is,

Roman,

with

more or

in

restricted

less

other

the Byzantine, and

comparatively

grouped

hand,

slight

the

and

while

fields,

in

Gothic,

superficial

styles,

extendvariations,

throughout the entire world influenced by a kindred

Hence

zation.

not

it

the

civili-

follows that reliefs or wall paintings can-

provide us with so direct and

respect to

age of

the

so valid

monuments

with

arguments
which

in

they

are connected, as the plan and the constructive details of the


architectural framework.

These conclusions are of especial force

at Assos,

where the
works

sculptures are extraordinarily provincial in character,

of different hands, displaying the greatest dissimilarity of con-

ception and execution

as, for

instance, the horse- and

legged centaurs of adjoining panels,


tectural features, of

oped

style,

yet where

human-

the archi-

a most pronounced and regularly devel-

upon the contemporary

are directly dependent

advances of the Doric peripteros

in

European Greece.

Hence

the synthetical methods of architectural history here find a

most direct and most trustworthy application.


This

will

become evident from a comparison of the temple

of Assos with

its

Towards the

immediate prototypes.
close

of the

sixth

century, before

the

in-

ARCH^OLOGICAL INSTITUTE,

300
of

terruptlon

the development of archaic Greek art by the

inroads of the Persians, architects of Attica and Aigina had


effected signal

improvements

in the

arrangement and propor-

tions of the Doric peripteros.

Recent researches have shown that the excessive elongation

had

Corinth,!

been

greatly

upon Cape Sounion,^ an


unfinished

The

at

a third

its

width

was

at

were

fifteen

older
to

temple

have been

demolition by the barbarians.

which was
Sounion

Corinth twice

at

less

than twice and

more

might be conceived from the bare


of the

Thus, while

oblong.

columns upon the side

The

both edifices being hexastyle.

the temple were at the


fifths

temple of

of

the

temple

there were at Sounion but thirteen, the fronts

of Corinth,
of

the

which appears

statement of the proportions


there

the

in

in

the appearance of the edifice being

thereby than

affected

its

stylobate,

half its width,^

reduced

edifice

the time of

of the

length

and a

noticeable

the archaic plan, so

of

absolute dimensions of

same time reduc