rs ot tgt g.rtj)icological Institute of %\mtM. 
 
 A PROTO-IONIC CAPITAL 
 
 KROIVI THE; SITE OK 
 
 NEANDREIA. 
 
 By JOSEPH THACHER CLARKE. 
 
 m ll^STITUIE \li 
 [r of 
 
 ,i:rV AMERICA. J'O'i 
 18/9. 
 
 BALTIMORE: 
 
 PUBLISHED FOR THE INSTITUTE BY 
 
 THE AMERICAN JOURNAL OF ARCHAEOLOGY. 
 1886. 
 
5Pn^trs of tlje ^rcIj;toIogicaI Institute of America. 
 
 ♦ 
 
 A PROTO-IONIC CAPITAL 
 
 KROM THK site: OK 
 
 NEANDREIA. 
 
 By JOSEPH THACHER CLARKE. 
 
 BALTIMOKE: 
 
 PUBLISHED FOR THE INSTITUTE BY 
 
 THE AMERICAN JOURNAL OF ARCHEOLOGY. 
 1886. 
 
JOHN MURPHY & CO., PRINTERS, 
 BALTIMORE. 
 
A PROTO-IONIC CAPITAL FROM THE SITE OF 
 NE ANDREI A. 
 
 Fig. 1. — Present condition of the block. 
 
 I. 
 
 This capital, — the most primitive memorial of the Greek Ionic 
 style as yet brought to light, — was found by the writer, Sept. 24, 
 1882, upon the summit of Mount Chigri, in the Troad. Chigri is 
 midway between Assos and Ilion, opposite Tenedos, and ten kilo- 
 metres from the coast of the Aegean. The extensive ruins upon the 
 site are, as will be shown, in all probability those of the ancient 
 Neandreia. They have never been disturbed by excavations, and for 
 more than 2,000 years this remote and precipitous height has been 
 uninhabited. During previous surveys, in 1881 and the spring of 
 1882, no sculptured stones or architectural members were to be seen 
 above the surface of the ground. But in the summer of the latter 
 
 3 • 
 
4 
 
 A PROTO-IONIC CAPITAL 
 
 year Turkish masons from the neighboring village of Yailadjyq, in 
 search of squared building-stones, had dug a shallow trench within 
 the city enclosure, exposing a corner of this block, which escaped 
 destruction because of its irregular shape. It was easily freed from 
 the soil, and was afterwards removed by Mr. Frank Calvert to the 
 farm of Akchi-Kieui (Thymbra), where it was carefully examined 
 and drawn. Together with it were discovered various fragments of 
 archaic terra-cotta, — portions of a leaved kyma, decorated with a 
 dark purple and black glaze like that found upon the most ancient 
 terra-cottas of Sicily. 
 
 The stone is a fine-grained volcanic tufa, of a light reddish-gray 
 color, obtained from a formation occurring in various parts of the 
 western and southern Troad. At Assos this material is employed 
 only in the oldest works, such as the lion's head which formed one 
 of the gargoyles of the chief temple,^ and a scroll believed to be part 
 of an akroterion of the same building. Tufa is never found among 
 later remains, and thus bears the same relation to the archaic archi- 
 tecture of the Troad as poros does to that of the Peloponnesos and 
 Sicily. The first Greek stone-cutters required a material more easily 
 worked than andesite, or even marble, and made up for the roughness 
 of the stone by priming the surface with stucco and painting it with 
 body color. 
 
 The capital remains in a state of preservation so good, that no 
 doubt can exist concerning any detail of the design. Some of the 
 corners have been split olF, nearly half of one of the volutes being 
 missing ; but in view of the friable nature of the tufa, and its long 
 exposure to the weather, the sharpness of the remaining tooled edges 
 is surprising {fig. 1). The building to which the capital belonged 
 must have been a ruin twenty-two centuries ago, and the block, when 
 found, was not protected by any great depth of earth ; yet the surface 
 has not been at all aifected by a decomposition like that which has 
 so obliterated many of the sculptures and mouldings of the harder 
 and coarser stone used at Assos. 
 
 The excellence of the design can have resulted only from an 
 acquaintance with many spiral prototypes ; and the admirable char- 
 acter of the technical execution is proof of a long practice in the 
 
 ^Now in the Boston Museum of Fine Arts: No. S. 1162. Cf. the writer's Report 
 on the investigations at Assos, 1881. Boston, 1882, p. 94, pi. 12. 
 
FR03f THE SITE OF NEANDREIA. 
 
 5 
 
 carving of similar details. The capitals of the later ages of Greek 
 art are of a higher and more organic development, better serving in 
 aesthetic respects as functional members of the columnar system ; but 
 they are rarely of better proportion, or of a more firm and graceful 
 outline {fig, 2). Too much emphasis cannot be laid upon the fact, 
 
 Fig. 2. — .Restored view: plan and section of capital. 
 
 thus evident, that this capital is by no means a first experiment in 
 the application of sj^iral forms to the upper member of a column, but 
 is rather to be considered as a link in the long chain of architectural 
 development which gradually led to the perfect forms of the capitals 
 of the Erechtheion. 
 
6 
 
 A PBO TO-IONIC CAPITAL 
 
 The helix is exact, and seems to have been determined by unwind- 
 ing a cord, to the free end of which was attached a chisel-point, 
 from a cylinder about 0.03m. in diameter, or perhaps, — for so great 
 a refinement is not inconsistent with the character of the desi<rn, — 
 from a slightly diminished cone as the evolute, fixed in the centre of 
 the perforation. The bordering fillets of the spiral vary in width 
 from 17mni. to 3mm., and are perfectly accurate to their very termi- 
 nation. The intelligent skill of the designer is especially to be seen 
 in the manner in which the leaves of the anthemion have been 
 profiled : their plane-angular, fluted, reeded, and concave-angular 
 sections securing a play of light and shade such as no geometrical 
 drawing can indicate (section AA, fig. 2). The incisions which sepa- 
 rate the surfaces of the volutes are deepened as they retreat from the 
 centre, gradually increasing from a shallow notch to a cut not 
 less than 0.11m. deep. The spiral line thus varies in appearance 
 from a light grey to a perfectly black shadow. The circular perfora- 
 tion in the centre of the volute, — the dfdalfio^ of the Erechtheion 
 inscription,^ — measures 0.125m. in diameter. It probably served 
 for the insertion of disks of some brilliant material, such as col- 
 ored marble, glass, or metal. This method of decoration had been 
 common in the Oriental prototypes from which the most charac- 
 teristic features of the Ionic style were derived, and, though seldoQi 
 adopted by the Greeks of a later period, was still employed in the 
 volutes of the fully-developed Ionic capital,^ as well as in the eyes 
 of the parotides and guilloche mouldings. The hole is cut com- 
 pletely through the stone, for what purpose is not clear. 
 
 The capital, at its point of juncture with the shaft beneath, is 
 not exactly circular in plan; the diameter from side to side being 
 0.01m. greater than from front to back. The summit of the 
 shaft must consequently have been slightly elliptical. This irregu- 
 larity of the stone-cutting is very remarkable in view of the perfec- 
 tion of the spirals and mouldings ; and, as the excess is in the axis of 
 the epistyle, it may have resulted from the capital, or more probably 
 the shaft, having been cut from a block not sufficiently thick to 
 allow one of the dimensions to equal the diameter determined by the 
 
 ^11.42. Hence termed oeulus by Vitruvius (iii. 5, 6), whose technical terms 
 are, for the greater part, translated from the Greek. 
 
 ^ As for instance in the Erechtheion, in the great temple of Ephesos, in that of 
 Sardis, etc. 
 
FROM THE SITE OF NEAJSDREIA. 
 
 7 
 
 designer. The capital was attacliecl to the drum adjoining it by a 
 cylindrical dowel, the hole for which (li, in plan fig. 2), 0.02ni. in 
 diameter and 0.055m. deep, is bored with great nicety. This pin 
 must have served as an axis for the grinding of the capital upon the 
 subjacent stone, during the last rubbing down of the bed surfaces. 
 The top of the capital, which is tooled to a perfect pkme, shows no 
 traces of dowels or clamps. The reverse of the stone is, in all the 
 main features of the design, the same as the front, but the details are 
 somewiiat less elaborate and the execution less careful. The scroll of 
 the back is slightly rounded in profile, but has no bordering fillets, 
 while the anthemion leaves are of simpler section, and without rims. 
 
 It is a question of much importance whether the shaft, to which 
 the capital belonged, was placed close to a wall as a stele, or was 
 employed as a constructive support in a building. The small dia- 
 meter of the column, and, especially, the fact that one side of the cap- 
 ital was evidently not exposed to close inspection, seemed at first to 
 favor the former view. After careful examination, however, the 
 writer became convinced that the capital surmounted a tall column, 
 probably standing in antis and supporting a wooden epistyle. 
 
 Notwithstanding the great projection of the volutes, — the width 
 of which far exceeds that customary in the steles of Greece, — the 
 bearing of the imposed weight is limited to the middle leaves of the 
 anthemion. If the block had been the capital of a stele, intended, 
 for instance, as a stand for inscribed stones or votive offerings, advan- 
 tage would naturally have been taken of the console-like projection 
 of the scrolls by a bearing upon the outermost leaves. This restric- 
 tion of the abacus to a surface less than half as broad as the capital 
 itself must have been determined by the consideration that, otlier- 
 wise, the slightest sagging of the epistyle-beam would have crushed 
 the sides of the volutes. From the extreme care taken to disengage 
 the outermost leaves of the anthemion from contact with the lintel, 
 it is evident that this precaution was held in mind. 
 
 The great projection of the volutes, as well as their shape, was 
 derived from traditional models. The form, originally determined 
 by the exigencies of a timbered construction, was here retained as a 
 mere decoration, filling out the corners between the vertical support 
 and the horizontal lintel. Thus, all the leaves of the anthemion and 
 the backs of the volutes approach very nearly to the soffit of the 
 epistyle, which, in the most closely related prototype {fig. 7), they 
 
8 
 
 A PROTO-IONIC CAPITAL 
 
 had actually adjoined. That the precaution was taken to restrict the 
 weight of the entablature to a part of the capital but little larger 
 than the upper diameter of the shaft, proves it to have formed part 
 of a constructive framework. The lack of dowellings between this 
 support and the imposed mass is, so far as it goes, in favor of the same 
 conclusion. The stones of Greek steles, because of their liability to 
 be displaced by lateral pressure, were commonly joined together by 
 metallic fastenings cast in lead ; but, for evident reasons, the abacus 
 of a true column is not often thus connected Avith the lintel above it. 
 
 The most conclusive argument, however, is to be derived from the 
 size of the block. A calculation based upon the proportions of mon- 
 uments of the fully-developed Ionic style leads to the assumption, 
 that the shaft and base belonging to the capital would, together with 
 it, reach a height of between four and four and a half metres. Even 
 this is considerably more than the height of the columns of many 
 prostyle temples ; and a building with columns in antis must neces- 
 sarily be assumed to have been of modest dimensions, especially in 
 the Troad.^ But it is probable that the actual size of the shaft was 
 greater than we should be led to expect from such a comparison. 
 The columns of primitive Greek architecture were, in general, more 
 diminished than those of the perfected styles ; the ratio of the upper 
 diameter to the lower, and to the height of the shaft, would conse- 
 quently have been smaller than that assumed. The fact that the 
 back of the capital is not treated with the same elaboration and care 
 as the front is explained by the assumption that it was situated at 
 some height, in a dark and narrow pronaos in antis, so that a good 
 view of the inner side could not be obtained. 
 
 All these points, — the excessive projection of the volutes, resem- 
 bling the original wooden prototype of the console-capital, the pre- 
 cautions taken to prevent the edges from being injured by a sagging 
 of the epistyle-beam, the fragile nature of the stone, and the small 
 diameter of the shaft, — lend weight to the supposition that the 
 entablature was formed, not of blocks of stone, but of timbers and 
 joists, such as those imitated in the fascias and dentils of the later 
 Ionic style. 
 
 The width of the capital is exactly twice its height, the volutes 
 being drawn in squares the sides of which are, as nearly as could be 
 
 * The Heroon of Assos, a Doric prostyles, has columns 3.6m. high. 
 
FROM THE SITE OF NEANDREIA. 
 
 9 
 
 measured M^thout instruments of precision, 0.594m. long (1 ft. llf 
 ins.). This dimension may with much probability be supposed to 
 equal two feet of the measure used by the designer, the result being a 
 unit of 0.297m. {t\ of an inch less than one English foot). The 
 thickness of the block is 0.357m., three-fifths of its height, or three- 
 tenths of the assumed unit. The question of the metrological 
 importance of these dimensions, and the decimal division of the foot 
 employed throughout the Troad in the earliest historical ages, should 
 not be entered upon until the stone has been measured with micro- 
 metrical exactness. 
 
 The exceptional interest of our capital lies in its historical signifi- 
 cance. It is one of the few memorials of the earliest period of 
 architectural development among the Greeks that have not been 
 swept away in the construction of the noble buildings erected soon 
 after the Persian wars, or of the showy edifices of the Diadochi. The 
 history of Neandreia will explain the circumstances which secured 
 the preservation of those remains not easily removed from the site 
 during ancient times. The capital cannot be ascribed to a later date 
 than the sixth century b. c. It is one of the many experiments 
 made by the Greeks of Asia Minor to determine the forms which, 
 according to the tradition preserved by Pliny ^ and Yitruvius (iv. 
 1, 7), were first employed in connection with a peripteral plan in the 
 primitive temple of Ephesos. 
 
 The testimony of antiquity is unanimous in the assertion, that the 
 Ionic style, as its name signifies, was derived by the European 
 Greeks from the eastern coast of the Aegean.^ Discoveries of the 
 present age have, further, made it evident, that the most characteristic 
 features of this style passed through the earliest stages of develop- 
 ment, neither in Greece nor in Asia Minor, but in Mesopotamia. 
 
 Before the application of the historic method to the study of the 
 derivation of architectural forms, the determination of the influences 
 which led to the adoption of the Ionic details was nothing but hope- 
 
 ^ Pliny's words (xxxvi. 56), in Ephesia Dianae aede primwn columnis spirae subdi- 
 tae et capitula addita, can only be referred to capitals, as well as to bases, of the Ionic 
 style. 
 
 ^Nothing can be more at fault than Boetticher's statement {Tektonik der Hellenen. 
 Berlin, 1874, second edition, vol. i. p. 165) that the Ionic style originated in Attica. 
 The entire position of this remarkable work in regard to questions of architectural 
 history is a warning against the misleading influence of aesthetic theories. 
 
10 
 
 A PROTO-IONIC CAPITAL 
 
 less conjecture. It is not long since, that scholars literally believed, 
 or at all events seriously considered, the explanation of the origin 
 of the style given by Yitruvius (iv. 1, 7), who relates that the Ionic 
 column imitated the proportions of a woman, — the volutes of the 
 capital representing the curled locks of hair; the flutes and fillets 
 of the shaft, the folds of the wide draperies ; and the base, the san- 
 dals. Thiersch,^ who occupied a position of the highest eminence 
 among classical scholars during the first half of the present century, 
 gave this picture a touch of reality by his identification of the Ionic 
 woman as a priestess with curled tainias tied about her ears. What- 
 ever may be the truth of the Vitruvian simile, as characterizing the 
 lightness and grace of the Ionic in comparison with the virile pro- 
 portions of the Doric, modern Avriters, in following the example of 
 the Roman maestro muratore, have not restricted their comparisons to 
 such pretty themes. 
 
 Winckelmann ^ suggested that coiled snakes may have served as 
 models for the volutes. Stackelberg^ argued that the twisted horns 
 of rams, suspended on the walls of primitive sanctuaries, or on the 
 corners of altars, were imitated by the original designer of the Ionic 
 capital. This idea was elaborated by Raoul-Rochette,^° and particu- 
 larly by Carelli,^' passing into the text-books through K. O. Mueller.^^ 
 Wollf believed that the bark of trees, placed upon the top of the 
 Doric echinos " before it had an abacus,'^ by curling round the block 
 had provided the starting point for the helix; while Hah n took 
 the spirals of marine shells as his model. Among the advocates of 
 such absurd prototypes we may note no less an authority than 
 
 'F. W. Thiersch, Ueber die Epochen der bildenden Kunst unter den Griechen. 
 (Second eilition) Halberstadt, 1829. 
 
 * J. J. Winckelmann, Versuch einer AUegorie, besonders fuer die Kunst. Dresden, 
 1766.' 
 
 ^ O. M. von Stackelberg, Der Apollotempel zu Bassae. Frankfurt am Main, 1826. 
 
 D. Raoul-Rochette, Monuments inedits d^ Antiquite. Paris, 1834. 
 
 F. Carellius, Dissertazione esegetica intorno aW origine ed al sistema della sacra 
 architettura presso i Greci. Napoli, 1831. 
 
 K. O. Mueller, Handbuch der Archaeologie der Kunst. (Third edition) Stuttgart, 
 1878, 54, 3. 
 
 J. H. Wolff, Aesthetik der Baukunst. Leipzig, 1834. This explanation has been 
 reiterated, during the past year, by H. Jennings, Phallicism, celestial and terrestrial, 
 heathen and Christian. London, 1884. 
 
 " G. von Halm, Motive der lonischen Saeule. Wien, 1862. 
 
FROM THE SITE OF NEANDREIA. 
 
 11 
 
 Viollet-le-Duc,^^ who conceived the Ionic vohite to have been copied 
 from curled sliavings left by the primitive carpenters upon the sides 
 of their wooden posts, illustrating this tasteless theory by a cut that 
 shows forms which wood could not assume under any treatment. 
 Even less satisfactory are those conceptions of an idealized spring, 
 taking the shape of an elastic cushion, which, placed upon the Doric 
 capital in the direction of the epistyle, is supposed to have been 
 squeezed out by the superimposed weight of the entablature so as to 
 curl again around the edges of the echinos. Chief among the pro- 
 fessors of this view is Guhl.^*^ This list mio^ht be ^reatlv extended. 
 Marini^^ gives the names of no less than twenty-six writers upon the 
 Ionic capital previous to the publication of his own work in 1825. 
 Some of the early treatises, such as those of Selva and De Rossi,^^ 
 display an ingenuity and a learning worthy of a better cause. 
 
 All these labored explanations of the significance and derivation 
 of the Ionic capital have fallen to the ground, — all this misdirected 
 antiquarianism has become a fit subject for ridicule, — upon the recog- 
 nition of the fact that a capital of anthemions and volutes, essentially 
 of the same character as that of the Ionic style, was customary in 
 Mesopotamia for centuries previous to the development of Greek 
 architecture, and is to be traced through Kappadokia, Phrygia, and 
 Phoenicia, to the coast of Asia Minor occupied by the Hellenes. A 
 great variety of terminal ornaments were formed by the designers of 
 Assyria in imitation of the radial leaves of the palmetto. The ends 
 of quivers, the plumes of horses' trappings, and other un^veighted 
 tips, appear of precisely the same shape as the conventional repre- 
 sentations of palm-trees upon Mesopotamian reliefs. When these 
 palmettos were so bound together as to form the so-called Tree of 
 Life, or such branches of flowers as are held by certain deities, the ends 
 of the connecting ribbons or the bracts were curled at the base, taking 
 the place of the bunches of dates seen under the palm-trees of the 
 reliefs. In architectural details this form Avas adopted, almost with- 
 out change, for the apex of steles. Among the ruins of the palace 
 
 ^^E. E. Viollet-le-Dac, Entretiens sur V architecture. Paris, 1858-72. Fig. 6. 
 ^''E. Guhl, Versuchueber das lonische Kapitael. Berlin, 1845. 
 L. Marini, Sul ritrovamento da me fatto deW metodo di descrivere la valuta loniea 
 Vitruviana, in the Atti dell'Accademia Romana di Archeologia. Eoma, 1825, vol. ir. 
 *®G. Selva, Dissertazione sulla valuta lonica. Padova, 1814. 
 ^^G. de Rossi, Esercitazione sulla valuta del capitello lonico. Firenze, 1817. 
 
12 
 
 A PROTO-IONIC CAPITAL 
 
 of Khorsabad^ a square ])ost has been preserved, in all respects like 
 the anthemion steles of Greece, the terminating palmetto being 
 the same as that continually occurring upon Greek vases and the 
 antefixes of early Greek temples. 
 
 It is with the higher development of these forms, through their 
 connection with the functional capital, that we are at present con- 
 cerned. By the adoption of the palmetto as an ornament inter- 
 mediate between a support and an imposed weight, the spread of 
 the leaves was necessarily much restricted. An increased importance 
 was thus assigned to the projections adjoining the shaft. It was 
 natural that this should have been made in the shape of a volute. 
 The spiral was, in every Avay, the form most pleasing to the early 
 Mesopotamian decorators. Not only did the ends of bows, the hilts 
 of swords, the carved ornaments of furniture, and the embroidered 
 
 BAG 
 
 Fig. 3. — Ivory -carvings from the north-western palace of Nimroud. 
 
 trimmings of robes assume this shape, but the spiral served in the 
 pictorial art of Assyria to represent objects really of entirely differ- 
 ent outline, such as entwined stems and leaves of plants, curls of the 
 human hair and beard, and even ripples of water. In short, the 
 spiral was as universal in the designs of Mesopotamia as were the 
 triangle and the zig-zag in those of Egypt. 
 
 Out of the ornamental spirals and palmettos of Assyria were 
 gradually developed the volutes and the anthemion of the Ionic 
 style. Semper, — most suggestive of writers upon the architectural 
 forms of the ancients, — displayed the extraordinary intuition for 
 
 20 V. Place, Ninive et V Assyrie. Paris, 1867-70, vol. iii. pi. 34. 
 
 One among many : Attic lekythos, with a representation of Orestes at the 
 tomb of Agamemnon, from the collection of Count Pourtales-Gorgier, published by 
 Raoul-Rochette in his Monuments inedits d'Antiquite, pi. xxxi. A. ; and also by 
 A. Maisonneuve, in his Introduction d V etude des vases antiques. Paris, 1817, pi. xxx. 
 
FROM THE SITE OF NE ANDREI A. 
 
 13 
 
 which he is remarkable, when he declared the evolution of the 
 Ionic capital to exhibit a stufenweisc Umbildung des zuerst nur eine 
 leiehte Palmette tragenden Volutenkelches in den balkenbelasteten Saeulen- 
 knavf.^'^ The capabilities of this combination for conventionalized 
 development led to its frequent employment in the details of various 
 architectural decorations. Several ivory-carvings from pieces of fur- 
 niture, found in tlie north-western palace of ^'imroud and now in the 
 British Museum, clearly show the Assyrian form of this capitaP^ 
 (^^.3). That marked A is not, strictly speaking, an architectural 
 detail. A latetal connection, visible beneath one of the volutes, 
 shows it to have formed part of such ornamental foliage as that 
 before referred to. The palmetto is consequently predominant and 
 of a semi-circular outline. B and c, on the other hand, show the 
 form as adapted to a functional capital. The leaves have decreased 
 in size and elaboration ; they have 
 become a simple anthem ion, and are 
 terminated by the straight line of 
 an epistyle. The volutes occup)^ 
 three quarters of the height; they 
 are of more independent formation 
 and better proportion. The hori- 
 zontal lines at the base are multi- 
 plied and emphasized, forming a 
 division between the capital and 
 the shaft similar to the annulets of 
 the floral columns of Egypt. The absolute similarity between these 
 two examples proves that the shape was a definitely determined type 
 of decoration. There is every reason to believe that these ivories are 
 exact representations of a capital systematically employed in Assyrian 
 architecture. They are essentially the same as the early Greek capital 
 of Mount Chigri, from which they diifer only in the imperfect spiral 
 
 G. Semper, Der Stil in den teehnischen unci tektonischen Kuensten. (Second edition) 
 Muenchen, 1878. Compare, also, J. Braun (Geschichte der Kunst, Wiesbaden, 1856- 
 58), "Der lonische SlU gehoert Niniveh, vielleicht hercits Babylon an; denn er ist der 
 gemeinsame Stil Asiens schon in unbereehenbar alter Zeit — er ist ein maechtiger Stil, dessen 
 Sendboten wir durch ganz Kleinasien und ueber die phoenikische Kueste nach KartJiago und 
 ins inner ste Afrika verfolgen koennenJ' 
 
 That marked A is shown in a small and inexact vignette serving as the tail- 
 piece to the list of illustrations in Layard, Discoveries in Nineveh and Babylon. Lon- 
 don, 1853. The others have not hitherto been published. 
 
 Fig. 4. — Supposed decoration of the 
 original wooden capital. 
 
14 
 
 A PROTO-IONIG CAPITAL 
 
 of the volute, and in the triangle masking the convergent lines at 
 the base. 
 
 Owing to the nature of the building-materials of Assyria, the 
 columns of that country were of wood,^* and but few vestiges of 
 them have been preserved. Fragments of wooden shafts, encased in 
 scales of bronze, have, however, been found,^'' and these suffice to 
 give certainty to the conclusions derived from the representations of 
 columns upon Mesopotamian reliefs. 
 
 No doubt can exist as to the origin of the spiral capital from the 
 application of the above-described forms to the details of such 
 wooden supports. As is the practice in every rational construc- 
 tion of timber, a horizontal block, projecting in the direction of 
 
 the imposed beam, was 
 placed between the sup- 
 port and the epistyle 
 {fig. 4). This interme- 
 diate member lent itself 
 readily to a decoration 
 of an I hem ion leaves and 
 lateral volutes. Spirals 
 were drawn upon the 
 projecting sides, either 
 in color or in incised 
 lines. Wooden columns 
 with capitals of this 
 kind, similar in design 
 to the ivory-carvings 
 before described, seem 
 to have been universal in Assyria, and to have formed, so to speak, 
 the only columnar order of the architecture of that country. They 
 appear in the well-known representations of sediculas, like that 
 
 Layard's workmen kindled their watch-fires with the timbers employed nearly 
 three thousand years ago in the construction of the palaces of the Assyrian kings. 
 Strabo (739), in an interesting passage relating to the buildings of Babylon, remarks 
 that Loth beams and columns were made of the trunks of palm-trees, the latter, 
 in the dwellings of the poorer classes, being wound around with twisted wisps of 
 straw, coated with stucco and painted. A more monumental method of this revet- 
 ment, referred to in Note 25, imitated the scales of the palm-tree in sheets of bronze. 
 
 A cylindrical column of cedar wood was found and published by Place, Ninive, 
 vol. I. p. 120, and vol. iii. pi. 73. 
 
 Fig. 5. — Assyrian cedicula from a relief. Khorsabad. 
 
FROM THE SITE OF NE ANDREI A. 
 
 15 
 
 standing in a royal park, upon a relief from the northern palace of 
 Koyundjik,26ancl that on the bank of a river, from Khorsabad'^ (fig. 
 5). The fact, that the spirals are, in these instances, so doubled 
 that four vohites appear between the shaft and the epistyle, does not 
 alfect the fundamental character of the ca])ita], this duplication being 
 due to tlie adoption of two transverse blocks of wood, instead of one. 
 
 Even more exact information concerning the appearance of the 
 Mesopotamian capital is to be obtained from the Sippara stone, dating 
 from about 900 b. c, now in the British Museum. Upon it is shown 
 one side of a tabernacle under which a deity sits enthroned, and it 
 is believed by Assyriologists that the artist has here imitated details 
 of the chief sanctuary of Sippara. The column is represented with 
 the greatest care. The slender shaft, evidently of wood, appears to be 
 covered, in imitation of the bark of a palm- 
 tree, with scales like those discovered by Place, 
 and the capital is of a spiral form, very similar 
 to the ivory-carvings (fig. 6). The volutes 
 spring from the shaft, fi'om which they are 
 separated by three annulets. They bear a bud 
 of semi-circular outline, of the same general 
 form as the authemion, and precisely like that 
 of a Phoenician capital found in Kypros ( fig. 7), 
 this abbreviation of the palmetto having been 
 rendered necessary by the cramped space be- 
 tween the scrolls, which did not allow an 
 indication of the separate leaves. The appearance of these details 
 upon the base, as well as upon the capital, of the Sippara column is 
 the clearest possible evidence of the timbered construction : the inter- 
 mediate block of carved wood being as much needed between the 
 base of the post and the sill, as between its summit and the epistyle 
 beam. The adoption of forms originally thus determined does not, 
 of course, disprove the assumption of Perrot,^^ that the capital of 
 
 A. H. Layard, Nineveh and its Remains. London, 1849, vol. ii. Published also 
 in G. Rawlinson, The Five Great Monarchies of the Ancient Eastern World. (Second 
 edition) London, 1871, vol. i. ; and in many other text-books. 
 
 2^Botta et Flandin, Monuments de Ninive. Paris, 1849-50, pi. 114. The illustra- 
 tion is taken from this work. 
 
 PervotetChiinez, Ilistoire del' Art dans UAntiquite. Vol. ii. Assyrie. Paris, 1883. 
 The author, however, certainly goes too far when, in the subsequent volume of the 
 
 Fig. 6. — Upper part of a 
 Babylonian column from 
 the Sippara stone. 
 
16 
 
 A PROTO-IONIC CAPITAL 
 
 Sippara was itself executed in sheet bronze, either soldered or 
 beaten to shape. The prototype, — the member to which the helix 
 and anthemion were first applied as a decoration, — was certainly of 
 wood. 
 
 The forms of the baluster, from its first appearance until the 
 present day, — during well-nigh three thousand years, — have never 
 overcome the one-sidedness resulting from this original timbered con- 
 struction. Contrary to the Doric and Corinthian capitals of the 
 ancients, to the trapeze-shaped capital of the Byzantines, and to the 
 cube capital of the Romanic style, the Ionic volutes, like the con- 
 soles of such Indian piers as those of the grotto of Ajanta, are 
 chiefly developed in the direction of the epistyle. When we, today, 
 employ the spiral capital, — whether placing the volutes vertically, 
 like those of Pompeii, or horizontally, according to Vignola's text- 
 book of the Renaissance, — we make use of forms which can be traced 
 back to the details of Mesopotamian ornament : in the same way 
 that so many of the words which we utter are derived, through many 
 transformations, from the primitive speech of our Indo-European 
 ancestors. 
 
 Long before the Greeks had built in the Ionic style, — while the 
 stone walls of the primitive fanes of Hellas still supported the beams 
 imitated in the Doric entablature, — the races inhabiting the plateaus 
 of Kappadokia, Lykaonia, and Phrygia, in Asia Minor, had derived 
 the chief features of their architecture from Mesopotamia. The spiral 
 capital of Assyria appears in Kappadokia in a city probably destroyed 
 as early as the time of Kroisos ; the Assyrian palmetto has recently 
 been found as the termination of a column in the great necropolis of 
 ancient Phrygia ; and the celebrated tombs of Lykia, especially 
 those of Antiphellos, Myra, and Telmessos, exhibit, together with 
 
 history, he assumes that the form of the volutes was suggested by a sphyrelaton 
 model. 
 
 Identified with Pterion in G. Perrot, Exploration areheologique de la Galatie et de 
 la Bithynie, d'une partie de la Mysie, de, la Cappadnce et du Pont ; executee en 1S61. 
 Paris, 1862-72. Compare a general view of the Ionic style given by the same author 
 in an essay on V Art de VAsie Mineure, ses origines, son influence, reprinted from vol. 
 XXV. of the Revue Areheologique in his Memoires d^ Archeologie, d Epigraphie et d'His- 
 toire (Paris, 1875); and in his short note Sur Vorigine de V Ordre ionique, in the 
 Bulletin de la ^;ociete Nationale des Antiquaires de France, annee 1871. 
 
 W. M. Ramsay, Some Phrygian Monuments. Reprinted from the Journal of 
 Hellenic studies, 1882, pi. xix. 
 
FROM THE SITE OF NEANDREIA. 
 
 17 
 
 late and debased forms, such primitive features as to place it beyond 
 doubt that this province was an important station in the advance of 
 the Ionic style from Mesopotamia to the Aegean. The most striking 
 examples of the intermediate stages of develoi)ment, however, are 
 the Phoenician works brought to light in Syria, Malta, and especially 
 in Kypros. All the varieties of the Assyrian volute are recognizable 
 among these remains. A capital discovered at Trapeza, near Fama- 
 gusta, Kypros, now in the Louvre,^^ {jig. 7), is of precisely the same 
 type as that represented by the ivories of Nimroud, and the capital 
 of Chigri. The clearest possible understanding of the development 
 of the Ionic volutes is gained by a comparison of this work with the 
 conventional decorations of Mesopotamia on the one hand, and the 
 
 Fig. 7. — Phoenician capital found at Trapeza, Kypros. 
 
 primitive Greek capital on the other, between which it forms a con- 
 necting link. The designers of Phoenicia, in adopting the forms of 
 Assyrian art, served rather to perpetuate than to perfect. In archi- 
 tectural history the importance of this conmiercial people consists in 
 their having spread abroad methods of artistic expression derived 
 by them from older civilizations, rather than in any great progress 
 of their own. The capital of Trapeza is an improvement upon the 
 Assyrian model, in that the volutes occupy the entire height between 
 
 ^' Perrot et Cliipiez, Hisfoire de I'Art dans V Antiquile. Vol. in. Pheiucie-Chypre. 
 Paris, 1885. The illustration is taken from this work. The height of this capital 
 is 0.75, the thickness 0.30, the length of the abacus 1.22m. 
 
18 
 
 A PBOTO-IOmC CAPITAL 
 
 the abacus and the annulets, and approach more closely to the true 
 curve of the helix. But in other respects the changes are few and 
 by no means advantageous. The square plan of the original wooden 
 post has been retained, together with the broad annulets and the 
 triangle of liard lines, which are connected with the base of the volutes 
 in a most inorganic fashion. The high abacus of long and narrow 
 plan, borrowed from the capitals of another style, restricts the devel- 
 opment of the spiral projections and cramps the anthemion to a mere 
 knop, of still less importance than that of the Sippara capital. 
 
 Even before the discovery of the capital from Mount Chigri, the 
 representations of Ionic details upon the most ancient Greek vases 
 made it evident that the primitive form of this member must have had 
 a much greater projection than that customary in the perfected exam- 
 ples, and that the volutes did not lie upon an echinos moulding, but 
 grew directly from the shaft, bearing between them an anthemion. 
 
 An archaic amphora from A^olcei, now 
 in the British Museum,^^ clearly shows 
 this formation (fig. 8). The painted and 
 incised outline upon this vase might be a 
 direct imitation of such a capital as that 
 
 ^ now brouo^ht to li2:ht, with which it a2:rees 
 
 Fig. 8. — Ionic capital from an . . ^ . . , . 
 
 archaic vase nven in proportions. Another archaic vase 
 
 of the same collection (No. 480) shows a 
 
 very similar capital. The architectural details of such paintings were 
 
 The painting on this vase is published by E. Gerhard (Auserlesene grieehisehe 
 Vasenbilder. Berlin, 1839-58, vol. iv. taf. ccxli) Lut without sufficient accuracy 
 for architectural comparison. Capitals of the kind are by no means uncommon: 
 compare F. Inghirami, Pitture di Vasi. Firenze, 1852-56. Vol. iii. Tav. 384. 
 
 In rare instances volute capitals of primitive form were executed in relief. One 
 of the most remarkable examples is the detail of terra-cotta, — possibly the handle 
 of a large vase, — found during the excavations at Assos. Its upright scrolls and 
 clumsy abacus are touched with white, the rest of the red clay being covered with a 
 dull-red surface-priming. This fragment is now in the collection of antiquities from 
 Assos, in the Boston Museum of Fine Arts, and is numbered P. 4121. 
 
 It is probable, also, that the heraldic sphinxes of Assos, upon the epistyle of the 
 chief temple of that place, rest their paws upon a diminutive Ionic stele, in the same 
 manner as the lions of the gate of Mykenai face an inverted proto-Doric column. 
 The surface of the stone has been so obliterated by weathering that the spirals can- 
 not be traced upon either of the reliefs; still, it is plain, from that in Boston num- 
 bered S. 1158, and especially from that shown by pi. 19 of the Assos Report, that the 
 sculptor has here represented the outline of the volutes, and even of the middle 
 anthemion, of an archaic Ionic capital. 
 
FROM THE SITE OF NEANDREIA. 
 
 19 
 
 declared by Hittorff to be les seuls souvenirs qui nous restent d' edifices 
 sacres sitr lesquels les notions historiques ont entib^ement disparu.^ 
 
 The discovery here published supplies the actual archaic example 
 hitherto wanting. 
 
 The Chigri capital shows a great improvement, even upon those 
 works most closely related to it which have been found in the inte- 
 rior of Asia Minor and in Kypros. Hellenic genius, brought to 
 bear upon this architectural member, at once manifests its superiority 
 in technical respects. The deformed volutes of the Kypriote capital 
 have become perfect spirals, while the anthemion leaves, as before 
 explained, are of great subtlety of design. The disturbing triangle 
 at the base of the volutes has been omitted; and the annulets, if 
 still employed, are transferred to the shaft. A decided progress is 
 noticeable in the omission of the Phoenician abacus, and the restric- 
 tion of the epistyle-bearing to a small part of the capital. In this 
 point the style has here already attained to its final perfection. 
 
 In other regards, it is still far removed from the most primitive 
 Ionic capitals of Greece hitherto known. The excessive projection, 
 considerably greater upon either side than the diameter of the shaft, 
 betrays the close dependence of the capital upon the forms of a 
 wooden construction. In striking contrast to the square termination 
 of the Doric column, the plan of this member is so oblong that the 
 front is more than three times as long as the side. A lateral projec- 
 tion so disproportionate, not only could never have originated in the 
 stone terminations of a round shaft, but it would be difficult to 
 believe from the drawing {plan, fig. 2) that the capital was executed 
 in that material. 
 
 So one-sided a capital could only have been employed in antis. 
 With the subsequent introduction of the peripteral plan it became 
 necessary to render the proportions of front and side more nearly 
 equal, in order to adapt the volutes to the corner column, — which 
 always presented the chief difficulty of this style.^* The important 
 
 Hittorff et Zanth. Architecture antique de la Sidle. (Second publication) Paris, 
 1870. 
 
 34 The assertion of Lohde (die ArchiteJdonik cler Hellenen, Berlin, 1862; reprinted 
 in J. M. von Mauch, Die architektonischen Ordnungen der Griechen und Boemer. 
 Seventh edition, Berlin, 1875), that the forms of the Ionic style originated in 
 connection with the peripteral and dipteral (!) plan is utterly incorrect, and would 
 be beneath criticism, were it not that it is made in a popular text-book. 
 
20 
 
 A PBOTO-IONIC CAPITAL 
 
 combination peculiar to the perfected Ionic capital, the conjunction 
 of an echinos with the volutes, was one of the means chosen to effect 
 this end. The front of the capital from Chigri is as entirely with- 
 out projection, as is that of the hypothetical Avooden support given in 
 figure 4 to illustrate the first application of the helix to the termina- 
 tion of a column. In the Erechtheion, however, the length of the 
 capital in proportion to its depth is found, when compared with that 
 of Chigri, to have been reduced by very nearly one half, the ratio 
 of the baluster to the front of the volutes being about 4 to 7. 
 
 The impossibility of allowing the epistyle to rest upon any part 
 of such volutes as those of the Chigri capital, and the desire to em- 
 phasize the horizontal lines of the termination, led to a further 
 change of much significance, namely, the inversion of the scroll in 
 such a manner that the two spirals no longer proceeded from the 
 shaft, but were connected by a horizontal band, upon the back of 
 which rested the narrow abacus and the epistyle-beam. This 
 arrangement is unquestionably of great antiquity, appearing upon 
 the before-mentioned relief of Kappadokia and in archaic vase- 
 paintings from Kypros. It was destined to wholly supersede the 
 upright volutes. But, as in the Doric style some primitive feat- 
 ures were retained in the antae-capitals, so, even in the latest 
 periods of Greek architecture, the principle of the vertical volutes 
 continued to be employed in the capitals of pilasters, as for example 
 in those of the great temple of Miletos, and of that of Athena Polias 
 at Priene. With this change in the position of the volutes the 
 anthemion ceased to be a constituent member of the Ionic capital ; 
 yet so entirely had it been identified with the style, that it remained 
 persistently in use as a subordinate decoration : appearing not only 
 in antefixes, simas, and decorated bands, but in the inner corners of 
 the spirals, and in the Attic necking of the capital itself. In the 
 archaistic capital of the temple of Bassae, the anthemion even assumes 
 its original position between the two volutes in the middle of the face. 
 
 There is but a single example known to illustrate the stages of 
 development intervening between the capital from Chigri and those 
 of the peripteral Ionic temples, namely, the fragmentary capital from 
 the Heroon of Selinous, probably referable to the sixth century B. c. 
 {fig. 9). Unfortunately, so little remains of this, that it is not even 
 certain whether the volutes were vertical or horizontal ; probabilities 
 favor the assumption of the latter arrangement, but in this respect 
 
FROM THE SITE OF NEANDREIA. 
 
 21 
 
 no great weight can be attached to the restoration given by Hittorff.^ 
 The helix, though it has more numerous turns, is very similar in 
 general character to that of the Chigri capital. The relative thick- 
 ness of the member is, however, much greater, and it is especially 
 remarkable that the roll, although not contracted as in all later 
 balusters, has been decorated with a pattern of scales. . Apart from 
 the too numerous convolutions of the spiral, the most immature 
 feature of the design is the excessive projection of the abacus, the 
 edge of which is ornamented with an egg-and-dart moulding. From 
 this it appears that the change in the position of the volutes led, at 
 first, to an extension of the bearing. This was again reduced in 
 
 Fig. 9 — Fragment of an archaic Ionic capital from Selinous. 
 
 subsequent times, the front of the strip receiving the same carved 
 ornaments as the side. 
 
 It is worthy of note, as an evidence of the tentative methods of 
 this period of advance, that the Ionic capital was, as in the Heroon 
 of Selinous, often employed together with the Doric entablature of 
 triglyphs and metopes: the capital developed upon the tall palm- 
 shafts of Asia thus being combined with the entablature derived 
 from the wall-plates and beams of primitive Hellas. It is not 
 strange that, among the few remains of this earliest period, but one 
 
 35 J. I. HittorfF, Restitution du temple d' Empedocle d Selinonte (Paris, 1851, pi. vi.), 
 and the work before quoted on the ancient architecture of Sicily. The illustration 
 is taken from the latter publication. 
 
22 
 
 A PROTO-IONIC CAPITAL 
 
 monument of so imperfect an arrangement should have been pre- 
 served until the present day. But the number of examples furnished 
 by the paintings of archaic Greek vases may be taken as an indica- 
 tion that the forms of the volute-capital had come into general use 
 at a period when the Ionic zophoros and dentils had not been intro- 
 duced into Greek architecture, or, at least, had not been developed 
 into a system. 
 
 The same elements that formed the capitals of the Erechtheion 
 constituted the terminations of the weak and overladen shafts 
 of Persepolis ; the spirals and palmettos of semi-barbarous Meso- 
 potamian decorations were employed as architectural details by 
 the designers of Persia, as well as by those of Attica. Yet the 
 decadence evident in the architecture of Persia is contemporary with 
 the highest development of the Ionic style among the Greeks. 
 ISTo better illustration is possible of the truth, that growth, and not 
 invention, is the j)rinciple of all progress in ancient art. 
 
 The builders of the present age have to deal with a confusion of 
 decorative forms and constructive methods similar to that which 
 prevailed throughout the ancient world before the rise of Greek 
 architecture. Hence, the most direct and practical service of archae- 
 ology to architecture must consist in a historical elucidation of those 
 principles of artistic selection and evolution which were followed by 
 the Greeks in their progress toward the incomparable perfection of 
 Attic monuments. 
 
 ^^As, for instance, the archaic vase in the British Museum, No. 480, and that pub- 
 lished by Inghirarai, before referred to. Many others have been collected by Hittorff. 
 It may be assumed, with great probability, that the combination of the triglyph- 
 frieze with the Ionic capital, observable in such later structures as the tomb of 
 Theron at Akragas, the Tomb of Absalom near Jerusalem, and several rock-cut 
 fagades in the great necropolis of Kyrene, is due to a reminiscence of the primitive 
 employment of these features upon the same building. It will be observed that 
 these monuments of Sicily, Syria, and Northern Africa are, although late, decidedly 
 provincial, and hence might naturally be expected to preserve barbarous and imma- 
 ture traits which had wholly disappeared from the art of Greece itself. 
 
FROM THE SITE OF NEANDBEIA. 
 
 23 
 
 II. 
 
 Reference has already been made to the fact, that the monuments 
 of the earliest period of architectural development among the Greeks 
 were, with but few exceptions, lost to science through having been 
 replaced by buildings of the more advanced styles. The preservation 
 of so primitive a memorial as this capital is to be accounted for by 
 the position of Mount Chigri, and, especially, by the history of the 
 ancient city which occupied its summit. The identification of the 
 ruins is of direct value in connection with the archaeological con- 
 sideration, as it supplies a terminus ante quern for all discoveries made 
 upon the site thus fortunately spared. 
 
 Chigri-Dagh is formed by cliffs of granite, rising steeply to a 
 height determined by the surveyors of the English admiralty as 
 1648 ft., and by VirchoAv^^ as 499.9 met. The barometrical readings 
 of the Assos expedition served only to verify these estimates, the 
 difference between which is but about three metres. Chigri is thus 
 the most prominent landmark of the Troad, north of Saqa-Kioh 
 and west of the Skamandros. 
 
 The view from the summit is magnificent. Upon the north is the 
 plain of Ilion, divided by the silver line of the Skamandros ; beyond 
 are the Hellespont and the Thrakian Chersonesos. Every curve of 
 the western coast of the Troad can be traced. The imposing ruins of 
 Alexandreia Troas, to which town the primitive inhabitants of Chigri 
 were removed by Antigonos, are easily discerned, and opposite to the 
 half-submerged mole of this once populous metropolis, lies Tenedos, 
 which, in still more ancient times, had been colonized by Tennes, son 
 of the king of Kolonai and Neandreia (Chigri), — thereby receiving 
 the name by which it is known even to-day. North of Imbros, 
 
 ^'Admiralty map No. 1608. Entrance to the Dardanelles, surveyed by Spratt, 1840. 
 R. Virchow, Beitraege zur Landeskunde der Troas. Aus den Abhandl. d. Kgl. 
 Akademie der Wissenschaften zu Berlin 1879. Berlin, 1879. 
 
24 
 
 A PBOTO-IONIC CAPITAL 
 
 Samothrake, the mighty seat of Poseidon, rises precipitously from 
 the blue waters of the Aegean, and far beyond the low and hazy 
 hills of Lemnos, the setting sun outlines with wonderful distinct- 
 ness the conical peak of Athos,^^*" more than one hundred and seventy- 
 five kilometres distant: thus, the horizon is bordered by the sacred 
 sites of the Kabeirian mysteries and the holy mountain of Eastern 
 Christianity; while prominent in the foreground is the domed mosque 
 of Kemaly. To the east stretch the fertile plains of the Samonion, 
 once a territory of Chigri ; beyond are the majestic heights of Ida. 
 On the south, the violet crest of Mount Lepethymnos, in Lesbos, 
 rises above the volcanic ridge which borders the Adramyttian gulf. 
 The scene of the Iliad is spread out before the beholder like a 
 map. 
 
 The uneven summit of Chigri is fortified by extensive walls, of an 
 irregular rhomboidal })lan. The greatest length of the enclosure, 
 from east to west, may roughly be estimated as one kilometre, while 
 its greatest width is less than one third as much.^^ The ramparts are 
 of hewn stones, polygonal and square, dating to various periods 
 anterior to the fourth century B. c. They are skilfully planned to 
 profit by the natural advantages for defence of this rocky height, and, 
 being in an exceptionally good state of preservation throughout their 
 entire length, they are among the finest monuments of Greek military 
 engineering in Asia Minor. The city is approached from the north- 
 east by a grand causeway, paved with slabs of stone, and evidently 
 
 This spectacle, little less than marvellous in view of the great distance from 
 shore to shore, has been observed by the writer on many occasions: from Chigri, 
 from the coast between Alexandreia Troas and Lekton, and even from the much 
 more remote summit of Mount Ida. It has been referred to by several authorities. 
 Clare conspicitur Athos cum caelum est serenum, ex HeUesponto et Asiatico liiore, multo 
 autem darius ex Ida Monte, says Vossius in the observations {ad lib. ii. cap. 2) attached 
 to his edition of Mela, Hagae Comitis, 1658. 
 
 We are reminded of the saying of the ancients, repeated by many writers, that 
 the shadow of Athos was cast upon the market-place of Lemnos at noon, or (and this 
 was undoubtedly the original meaning of the fable) by the setting sun at the time of 
 the summer solstice. 
 
 Newton, whose work will be cited below, judges the summit to be ''more 
 than a mile long," from the fact that it took him twenty minutes to walk the dis- 
 tance; but it is evident that this estimate is too great. Calvert's measurement, 
 published by Pullan and repeated, without acknowledgment, by Schliemann, gives 
 1900 paces as the length, and 520 paces as the breadth of the enclosure. 
 
FROM THE SITE OF NEANDREIA. 
 
 25 
 
 of great antiquity. The chief entrances to the enclosure are at the 
 north-cast and at the south, and are particularly important. They 
 are flanked by square towers very similar to those of the main gate- 
 way at Assos, their monolithic lintels and jambs showing traces of 
 the bolts and battens. It is not the present purpose, however, to 
 give any adequate account of these fortifications, or of the ruins of 
 the city itself ; though it may be remarked that the capital which is 
 the subject of this paper and the fragments of the painted terracottas 
 which undoubtedly belonged to the same building, were found in the 
 north-west corner of the enclosure. 
 
 The first explorer who is known to have visited the ruins of 
 Mount Chigri is Pococke, — the earlier travellers in the Troad who 
 penetrated beyond the port of Alexandreia Troas, such as Bel on 
 (1554) and Du Loir (1654), not having gone farther inland than the 
 hot springs of Lidja. Pococke""^ calls the site Chigur, and identifies 
 it with Skepsis, from the similarity of the name of that ancient town 
 to that of the neighboring village of Eskiupjee (Eski Skupchu). De 
 Vaugondy's ancient map of Asia Minor," published fifteen years 
 after Pococke's last volume, gives Cocyllum (Kokylion) in the posi- 
 tion of Chigri. Kokylion is one of the towns of the Troad mentioned 
 by Pliny as deserted in his time, and its identification with Chigri 
 rests solely upon the similarity of the names. Whether this was 
 due to the map-maker alone, or to some traveller previous to 1760, 
 other than Pococke, it is not possible to say. Lechevalier^^ subse- 
 quently adopted the name Kokylion from the village of Qocholo- 
 bassy, to the north of Chigri, which mountain he calls Kiril-Dagh. 
 This misleading method of identification was also practised by 
 Choiseul,^^ whose assumption that Chigri was the site of Kenchreai 
 is still the most generally accepted. ChoiseuPs authority was in this 
 respect greatly strengthened by the endorsement of Leake and 
 
 ^"E. Pococke, A description of the East and some other countries. London, 1743-45. 
 Part two. 
 
 Asia 3Iinor. Auctore K. de Vaugondy. Paris? 1760? 
 ^2 Pliny, V. 32. Compare also Xenophon, Hell. iii. 1. 16. 
 J. B. Lechevalier, Voyage de la Troade, fait dans les annees 1785 et 17S6. (Third 
 edition) Paris, 1802. 
 
 M. G. A. F. de Clioiseul-Gouffier, Voyage pittoresqiie de la Grece. Paris, 1782- 
 
 1809, Vol. II. 
 
 *^ W. M. Leake, Journal of a Tour in Asia Minor. London, 1824. 
 
26 
 
 A PEOTO-IONIC CAPITAL 
 
 Webb/^ Some account of the interesting geological aspects of Chigri 
 is given by Tchihatchetf ; he makes, however, the error of speaking 
 of the formation as a trachyte. The more modern travellers who have 
 visited the ruins are ISTewton/^ whose excellent description has been 
 referred to ; Pullan/^ who published Calvert's notes ; and, within 
 the last few years, Meyer,^^ Schliemann,^^ Virchow,^^ Diller, the 
 geologist of the Assos expedition,^^ and Jebb.^* 
 
 Compared with the many visitors to the neighboring towns, this is 
 but a short list. Perhaps the neglect of Chigri may in some measure 
 be attributed to the evil repute of this lonely mountain as the resort 
 of brigands. Commander Spratt having had a narrow escape from 
 one of these bands while visiting the site. Many travellers have 
 passed directly by the foot of the hill on the road from Ezine to the 
 ruins of Alexandreia Troas, without making the ascent. 
 
 The identification of Chigri as Kenchreai, proposed by Choiseul 
 and favored by Leake, Webb and Virchow, is, as before mentioned, 
 
 P. B. Webb, Osservazioni intorno alio stato antico e presente delV agro Trojano ; 
 first published in Acerbi's Bihlioteca Italiana, Milnno, 1821; written by the author 
 for that journal and translated under his supervision. 
 
 P. Chikhacliev, Asie Mineure, description physique, statistique et archeologique de 
 cette contree. Quatrieme partie. Paris, 1853-69. 
 
 C. T. Newton, Travels and discoveries in the Levant. London, 1865. 
 
 R. P. Pullan, in Murray's Handbook for travellers in Turkey in Asia. (Fourth 
 edition) London, 1878. 
 
 E, Meyer, Geschichte von Troas. Leipzig, 1877. 
 ^' H. Schliemann, Ilios: Stadt und Land der Trojaner. Leipzig, 1881, The slight 
 notes given in the Beise in der Troas im Mai 1881 (Leipzig, 1881) are reprinted 
 in Troja. London, 1884. Scliliernann's statement {Ilios p. 57), that there is no 
 accumulation of debris on Mount Chigri, is misleading. The native rock does, indeed, 
 crop out in many parts of the fortress, notably at the south-east and north-east cor- 
 ners, where peaks of trachyte rise even above the fortification walls : yet, through- 
 out the greater part of the enclosure, there is a soil of considerable depth, as is evi- 
 dent from the fact that the summit of the mountain serves as the pasture for a great 
 number of horses and cattle at a season when the lower plains have been parched by 
 the summer sun. Schliemann's further assertion, that "only here and there a late 
 Roman potsherd and some fragments of bricks of a late date " were to be seen, is 
 absolutely incorrect. Careful examinations of the site, on several occasions, failed 
 to bring to light any remains more recent than of the fourth century b. c. 
 
 Virchow's barometrical measurement of the height, given in the Beitraege zur 
 Landeskunde der Troas, quoted above, is printed also in Schliemann's Ilios. 
 
 J. S. Diller, The geology of Assos, in Clarke's Beport on the investigations at Assos, 
 1881. Boston, 1882. 
 
 R. C. Jebb, A tour in the Troad : in the Fortnightly Bevieiv, No. cxcvi. London, 
 1883. 
 
FROM THE SITE OP NE ANDREI A. 
 
 27 
 
 that generally accepted.^^ This assumption can be definitely dis- 
 proved. Kenchreai is of interest as one of the cities which claimed 
 to have been the birthplace of Homer (Souidas, s. v. OfirjpoQ), and 
 as the place Avhere the great poet dwelt while familiarizing himself 
 with the scenes of the Trojan war (Steph. Byzant. s. v. Ksyj^fjiac). 
 But Kenchreai existed as a citadel at a date long after Chigri must 
 have been deserted. Georgios Pachymeres [De Mich. Pal. vi. 24) 
 informs us that the emperor Michael Palaeologos confined the 
 unfortunate Manuel in this fortress. The same writer [De Andron. 
 Pal. v. 27) describes in detail the taking of Kenchreai by the Turks, 
 
 Fig. 15. — Sketch-map of the ancient Troad. 
 
 soon after the beginning of the fourteenth century : he relates that, 
 after having held out for some time, it was compelled to surrender 
 from lack of water, and was burned by the enemy. Nothing is 
 more certain than that this citadel Avas not situated upon Mount 
 Chigri, where no Byzantine remains whatever are to be met with. 
 The writers who have advocated the identity of Kenchreai and 
 
 Compare : J. A. Cramer, A geographical and historical description of Asia Minor. 
 Oxford, 1832 ; and C. Texier, Asie Mineure, description geographique, historique, et 
 archeologique des provinces et des villes de la Chersonnese d' Asie. Paris, 1862. One of 
 the volumes of L' Univers. 
 
28 
 
 A PBOTO-IONIC CAPITAL 
 
 Chigri must either have been ignorant of the reference made to that 
 ancient town by Pachy meres, or not well acquainted with the char- 
 acter of the remains upon the site. Kenchreai is undoubtedly to be 
 identified with Kiz-Kalessi, — a citadel upon the north of Chigri, and 
 one of the few sites of the Troad which were fortified in Byzantine 
 times. Not having been occupied by the Turkish conquerors, it 
 still shows traces of the fire by which it was destroyed. 
 
 The ancient atlas of Smith, and that of Kiepert, as well as the 
 map in Mueller and Duebner's edition of Strabo, place Kolonai upon 
 the site of Chigri. In like manner Eduard Meyer, one of the best 
 informed of all the travellers in the Troad, speaks of the remains as 
 those of Kolonai. It is not strange that this commanding height 
 should have been identified with the stronghold chosen as a retreat 
 by the Spartan Pausanias while carrying on his treacherous negotia- 
 tions with the Persians.^*^ Nevertheless it is certain that Kolonai 
 was situated much nearer to the sea than Mount Chigri. Xeno- 
 phon {Hell. iii. 1, 13 and 16) twice mentions it as a maritime town, 
 and the testimony of Strabo is even more explicit, for he describes it 
 as lying on the sea (589), and on the coast opposite Tenedos (604). 
 The latter assertion is made also by Diodoios (v. 83. 1) and by Pau- 
 sanias (x. 14. 2). As will be explained below, the passage of Sky- 
 lax in which Kolonai is mentioned must be taken in the same sense. 
 Among those ancient writers whose mention conveys any indication 
 of the situation of the town, there remains only Pliny (v. 32), who 
 says distinctly enough intus Colone intercidit, but whose testimony 
 concerning the Troad is of but little value, especially in the case of 
 those cities which, like Kolonai, were deserted more than three centu- 
 ries before his time. Even the name KoXcouai is characteristic of 
 such mounds as those of the tertiary formation found on this coast 
 of the Troad, and would be entirely inexplicable in connection with 
 the granite mountain of Chigri.^^ 
 
 ^6Thouk. I. 131 ; Diod. xiv. 383; Corn. Nep., Pans. 3. 
 F. Calvert, — On the site and remains of Colonae, in the ArchcBological Journal, 
 vol. XVII. London, 1860, — believes the narrow summit of Beshik-Tepeh, three 
 miles north of Eski-Stambol (Alexandreia), to be the true site of Kolonai. But the 
 distance of this place from Strabo's Ilion is less than the one hundred and forty 
 stadia designated by the geographer. It ai)pears, moreover, from another passage 
 (Strabo, 604) that Alexandreia was founded between the tract known as the Achaiion 
 and Kolonai, and that we must consequently look for the latter town south of the 
 great metropolis of the Diadochi. 
 
FROM THE SITE OF NEANDREIA. 
 
 29 
 
 All indications favor the identification of Mount Chigri with the 
 ancient Neandreia. This view, first suggested by Calvert,^ is based 
 upon the description given by Strabo. Strabo states that the Nean- 
 dreians were situated above Hamaxitos, — the position of which town 
 is determined, by the notices of it in other passages, as close to Lek- 
 ton (604), near Larissa (440) and the Sminthion (605),— on this 
 side (i. e. to the north) of Lekton, but further inland and nearer 
 II ion, from which they were distant one hundred and thirty stadia.^® 
 Strabo states, also, that the territory of Assos and its colony Gar- 
 gara was bounded by the tracts belonging to Antandros, Kebrene, 
 Neandreia and Hamaxitos (606), towns which are thus seen to lie 
 almost in a semi-circle around the region in question ; and further, 
 that the plain of Samonion (now known as that of Bairamitch) 
 belonged to Neandreia (472), — a district that would naturally be under 
 the domination of the stronghold of Mount Chigri. Strabo more- 
 over tells us that the inhabitants of Neandreia, together with those 
 of many other cities of this region, were removed by Antigonos to 
 the newdy established town of Alexandreia Troas. Pliny, a little 
 later, speaks of the site as deserted (v. 32). 
 
 A similar conclusion is to be derived from a mention of Neandreia 
 by Xenophon (Hell. iii. 1, 13-16). Mania, the satrapess of the 
 province, whose chief seat was in the interior of Kebrene and Skep- 
 sis, possessed Neandreia, and extended her dominion by reducing the 
 maritime towns of the south-western Troad, Larissa, Hamaxitos and 
 Kolonai, which had remained in the possession of the sea-faring 
 Greeks. On the arrival of the Spartan Derkyllidas, these three 
 towns surrendered at once, as did, within one or two days, Nean- 
 dreia, Ilion, and Kokylion, after the fall of which places Kebrene 
 was besieged. Xeno^ihon's enumeration of the towns can leave no 
 doubt as to the route followed by Derkyllidas. The Spartan gen- 
 eral must have landed at the ancient port of Lekton, and have 
 moved into the valley of the Skamandros by the natural pass upon 
 the north of Mount Chigri, taking the town of Neandreia upon 
 its summit, which, it is to be observed, is the first named after leav- 
 ing the sea at Kolonai. At the present day, the main road of the 
 
 ^^F. Calvert, On the site and remains of Cebrene, in the Archceological Journal, vol. 
 XXII. London, 1865. 
 
 Strabo, 606. Korai's emendation, fieo aye tor spot dk for /xeaoyeioTepi re, is self-evi- 
 dent. 
 
30 
 
 A PROTO-IONIC CAPITAL 
 
 country, by which the wine of Tenedos is carried to Ezine and Bai- 
 ramitch on the Mender6, follows the same route. 
 
 Opposed to this weight of evidence, we have the statement of 
 Skylax (p. 36) that Neandreia was situated on the sea. As at least 
 those portions of Skylax relating to the coasts of Asia Minor are to 
 be referred to a date anterior to that of the foundation of Alexandreia 
 Troas and the depopulation of Neandreia, it would be natural to give 
 entire credence to this earlier authority, and to assume that Strabo, 
 although evidently quoting from Demetrios of Skepsis, was mistaken 
 in his identitication, — were it not that it is plain, from internal 
 evidence, that the passage in question is, as it stands, a misstatement 
 throughout. Skylax, whose Periplous was characterized even by 
 Bentley as one of the most corrupt books in the world," gives in 
 his description of the Troad two lists, the one of inland towns : 
 Sigeion, Achilleion, Achaiion, Kolonai, Larissa, Hamaxitos and 
 Chrysa, — the other of towns on the sea : Kebrene, Skepsis, Nean- 
 dreia and Pityeia. Now all those of the first list are well known to 
 be situated upon the coast, while, of the latter list, both Kebrene 
 and Skepsis were far inland. Pityeia does not belong to the Troad 
 at all. It is thus plain that the classifications of the towns have 
 been interchanged : that those of the first list were originally 
 described as situated on the sea, those of the latter as in the interior.^^ 
 
 ^^The words of Skylax are: 'Kal ev rfj rjizeipifi 'Elyr/ Kal 'Axi^?i-eiov Kal KpaTfjpe 
 'Axfiiojv, Ko?iO)vai, Adpiaaa, 'A/m^irbg Kal lepov ' ATcolTiuvoq^ Iva Xpvo?jg lepdro. 'Evrevdev 
 t)E Alo?ug x^P*^ na/ishai. AioMdeg de noXeig ev avrrf elatv eiri OaXdrrrj aide Kefipi^Vj XK7)iptg^ 
 'Nedvc^psca^ Uirveta. 
 
 The difficulty presented by this passage was evident to Mueller, and in a note to 
 his edition of Skylax (Geographi Graeci minores, Parisiis, 1855, vol. i.), he inserts 
 between aiJe and KejSpijv the words: "AnGog, Vdpyapa^ " AvravSpog' ev 6e fitaoyela aLde. 
 This empiric change of the sense is actually adoi)ted in the text of the last critical 
 edition of Skylax, Anonymi viilgo Scylacis Cai-yandensis periplum maris inlerni recensuit 
 B. Fdbricms (H. T. Dietrich) Lipsiae, 1878. It by no means meets the difficulties 
 of the case, the maritime towns still being described as inland. Were it desirable 
 to restore the text, it would be more reasonable to simply interchange the lists, and 
 not attempt to bring in the names of Assos, Gargara and Antandros. The towns on 
 the Gulf of Adramyttion would not have been named before Kebrene and Neandreia. 
 
 It is surprising that so manifest a corruption sliould have misled writers upon 
 ancient geography, otherwise most trustworthy. Thus, C. Mannert [Geographie der 
 Griechen und Boemer aus ihren Schriften dargestellt. Leipzig, Nuernberg, Landshut, 
 1829-31. Third edition) and A. Forbiger {Handbuch der alien Geographie. Leipzig, 
 1842-44) refuse all credence to Strabo, on the strength of this passage of Skylax. 
 The latter author, in his second volume, describes Neandreia as a maritime town, 
 
FROM THE SITE OF NEANDREIA. 
 
 31 
 
 Hence, the testimony of Skylax may even be claimed in support 
 of that of Strabo. 
 
 A passage of the greatest importance in reference to Neandreia, 
 and one to which attention has not been called by any writer upon 
 the geography of the Troad, is given in Dictys of Krete.*^^ From 
 this we learn that the Greeks before II ion, being harassed by attacks 
 of the inhabitants of the neighboring country, moved their forces 
 against the towns situated nearest to Ilion, first invading the realm 
 of King Kyknos, the chief place of which was Neandreia. The 
 Greeks took this citadel, and were about to destroy it by fire, but 
 were persuaded to spare it by the prayers and tears of the inhabitants, 
 who tendered their submission to the invaders and gave up the two 
 sons and the daughter of King Kyknos, he himself having been slain, 
 some time before, by Achilleus. Advancing, thereafter, beyond 
 Neandreia, the Greeks reduced Kylla, but left Kolonai unharmed, as 
 that town belonged to the Neandreians and was protected by the 
 alliance which had been concluded with them. 
 
 The manuscripts of Dictys, differing among themselves, show cor- 
 ruptions of the names : Neandreia appearing as Meandria, Mentore 
 or Metore, Kolonai as Corone. The first of these errors (3feandri- 
 
 "east of Gargara." Compare his position in Pauly ( Beal-Encyclopcedie der clas- 
 sischen Alierthumsivissenschaft, s. v. Neandria vol. v. Stuttgart, 1848), where he under- 
 stands Skylax to place the town on the Hellespont. Tliis is translated, with- 
 out acknowledgment, in the notice on Neandreia which, signed by Leonhard 
 Schmitz, appears in Smith's Dictionary of Greek and Roman Geography, London, 
 1873. Forbiger's erroneous quotation of Skylax in support of the statement 
 that Neandreia was on the Hellespont is thus perpetuated. It may be remarked 
 that Smith not infrequently presents to his readers stolen and garbled versions of 
 Pauly's articles. 
 
 ®' Dictys Cretensis, it. 12 and 13. The author twice refers to the realm of King 
 Kyknos as adjoining Ilion. 
 
 The interest of the passage in question is not restricted to the geographical indi- 
 cations which it affords; it also furnishes an argument in favor of the belief that, in 
 this much discussed work, there have been preserved, together with later and spuri- 
 ous material, some traditions of great age which are credible in the same sense as 
 are those collected in the Homeric poems. Though the events recorded should be 
 consi;lered as romance rather than as history, the geography could not thus be 
 invented. The author of the original work must have had an intimate acquaintance 
 with the Troad, or at all events must have derived his information from sources of 
 this character now lost to classical science. This may be well illustrated by a com- 
 parison of the work of Dictys with that of Dares, whose vapid descriptions of the 
 Homeric heroes contain no mention of geographical details, or do not differ in 
 these particulars from the earlier writings from which the book was compiled. 
 
32 
 
 A PROTO-IONIC CAPITAL 
 
 orum for Neandriorum, etc.) was pointed out nearly two centuries ago 
 by the learned Perizonius.^^ He based his conviction solely upon 
 the accounts of Kyknos given by Malala (p. 124, Oxford ed.) and 
 Kedrenos (p. 221), who, evidently deriving their information from 
 Dictys, assert that King Kyknos lived in Neandreia, near llion. It 
 is well known that these Byzantine writers frequently quote the ipsis- 
 sima verba of the Greek Dictys. They were in possession of the 
 original work, which has since disappeared, and their rendering of 
 the geographical names is hence far more worthy of confidence than 
 that of the copies of the Latin version of Dictys, now alone acces- 
 sible to us. The conjecture of Perizonius thus admits of no doubt, 
 and this correction is adopted in the latest critical texts. 
 
 The emendation Colonen for Coronen, naturally following the 
 Latin orthography of the name as given by Pliny (v. 32) was 
 suggested by Fuchs.''^ The three ancient writers who differ from 
 Dictys, Malala, and Kedrenos in the name of the capital of Kyknos, 
 namely Diodoros (v. 83. 1), Strabo (589 and 604) and Pausanias 
 (x. 14. 2), agree in speaking of Kolonai as his dwelling-place. It is 
 surprising that, notwithstanding this weight of argument, the emen- 
 dation has been refused by Dederich,^^ and is not even referred 
 to by Meister;^ their editions of Dictys, the most recent published, 
 vStill read Corone, while no place of that name exists in the Troad. 
 
 The testimony of the author of the Greek original must have been 
 founded upon traditions, oral or written, which show an accu- 
 rate acquaintance with the country around llion. Whether these 
 legends do or do not recount the actual events of a predatory war- 
 fare, carried on by the Achaians in the Troad, they must at least 
 have been so framed as to appear credible to the Greeks inhabiting 
 this remarkable country during the historic period. As it is now 
 read, by the aid of the Byzantine plagiai'ists and in the light of a 
 familiarity with the Trojan landscape, the passage describes occur- 
 rences w^hich would naturally have taken place in such a campaign. 
 
 J. Perizonius, Dissertatio de historia belli Trojani, etc. (Leyden ?), (1701 ?). This 
 essay was incorporated in the edition of Dictys published by L. Smids, Amsterdam, 
 1702, and in others since then. 
 
 ®^ J. A. Fuchs, Pe varietate fabularum Troiearum quaestiones. Coloniae ad Rhenum, 
 1830. This excellent work is but very little known ; the copy which has been on 
 the shelves of the British Museum for half a century was found to be uncut. 
 
 Dictys Cretensis, Belli Trojani libri sex. Ed. A. Dederich, Bonnae, 1833. 
 
 Dictys Cretensis, Ephemeridos belli Troiani. Ed. F. Meister, Lipsiae, 1872. 
 
FROM THE SITE OF NE ANDREI A. 
 
 33 
 
 According to the narrative of Dictys, the Greeks disembarked at the 
 mouth of the Mendere, near the modern Koum Kaleh, and encamped 
 in tlie plain. On their expedition against the country of King Kyknos 
 they passed up the valley of the river, through the defile of Bali- 
 Dagh, to the stronghold of Mount Chigri. After having come to 
 terms with the inhabitants, the Greeks found but two courses open 
 to them : to advance inland, across the plain of Bairamitch, or to 
 turn to the south-east, towards the coast. By a further inroad they 
 would have incurred the danger of being cut off by the enemy. 
 Undoubtedly influenced by this consideration, the Greeks chose the 
 latter alternative, reaching the sea south of Eski Stambol. 
 
 It is thus plain, that the legends of the Trojan cycle relating to 
 King Kyknos originally designated both Kolonai and Neandreia as 
 towns of his kingdom ; the former as a seaport, the latter as a 
 mountain fastness. All the episodes in the life of the hero relate to 
 the sea, and it is probable that his residence was Kolonai, as the 
 more trustworthy authorities assure us. The opposite island of 
 Tenedos was colonized and named after Tennes, a son of Kyknos,^^ 
 and the most prominent part taken by the king in the Trojan war 
 was an attempt to prevent the Greeks from lauding.^^ The citadel 
 of Neandreia, on the other hand, must have been a stronghold and 
 retreat ; this is sufficiently indicated by the tradition given by Dictys 
 (ii. 13), that it was the abode {nutrix) of the children of the king. 
 The legend which asserts Kyknos to have been the son of Poseidon 
 and Skamandrodike^^ must doubtless be taken as significant of the 
 se^ and the river which formed the boundaries of his realm. 
 
 It is worthy of note, in this connection, that both Xenophon {Hell. 
 III. 1. 16) and Strabo (472 and 606) speak of the Neandreians as a 
 people ; the latter, as has been seen, describing not the position of 
 the town, but that of the tract which bore its name. This is 
 explained by the fact that Neandreia, like Assos, was, at a very 
 early period, the capital of a small independent kingdom, which con- 
 tinued in the memories of the inhabitants long after the entire Troad 
 had been included in a much wider dominion. 
 
 ®^ Steph. Byzant. s. v. TeveSoc ; Suidas, s. v. levkdioQ avOporrog ; Cicero, in Verrem, 
 act. IT. 1. 19; Konon, Narrat. xxviii ; Plutarch, Quaest. Graec. 297; Servius, Com- 
 mentary to Virgil, Aen. ii. 21 ; and the other authors quoted in this connection. 
 Aristotle, RheL ii. 22. 12. 
 
 Scholiast to Homer, 11. A. 38; Scholiast to Pindar, 01. ii. 147; Tzetzes, ad 
 Licoph. 233 ; and Eudocia, Viol. p. 264 ; for a different account, see Hyginus, Fab. 157. 
 
34 
 
 A PROTO-IONIC CAPITAL 
 
 No further information is to be derived from the references to 
 Neandreia made by Theopompos (Fr. 310) and Charax (Fr. 4), 
 preserved in Stephanos of Byzantion.^^ The town was naturally 
 inckided in the province of the Hellespont, and was so described. 
 These writers are, however, to be quoted as completing the list of 
 ancient authors who make any mention of this place. 
 
 The indications obtainable from the coins of Neandreia lend 
 further weight to the conclusions derived from classic literature. 
 The head of Apollon, who was evidently the chief deity of the 
 town/*^ generally appears upon the obverse, Avhile the symbols of the 
 reverse suggest the advantages derived by the inhabitants of the 
 V citadel from the pastures and cultivated fields of the great Samonian 
 Plain. A grazing horse and an ear of wheat formed the most com- 
 mon types/^ and a ram occurs upon a fine coin of the fifth century, 
 now in the British Museum .^^ It is a point particularly wortliy of 
 remark, that a coin published by Sestini^^ was re-struck with an incuse 
 stamp AAEEAN, without doubt to facilitate the circulation of the old 
 mintage in the city to which the inhabitants had removed. It is 
 evident, that the Neandrians formed an influential part of the popula- 
 tion of Alexandreia Troas, from the fact that the coins of the latter 
 town, during the first three centuries of its existence, display the 
 two types characteristic of the Neandreian mintage : the head of 
 Apollon^* and the grazing horse. 
 
 ''^Steph. Byzant. s. v. '^edvdpeLa. Compare also the mention s. v. *avna. 
 It must have been with reference to this cult, and to that of the neighboring 
 Chrysa, that Strabo (p. 618) declared ApoUon to be the chief deity of the south- 
 western Troad. Tenedos also worshipped ApoUon, following in this respect its 
 parent city, even as early as the time of the Homeric poems [Iliad, i. 38, 451, etc.), 
 in the same manner as did Alexandreia Troas in the age of the Diadochi. The 
 especial protection granted by Apollon to the Trojans and their allies, and the preva- 
 lence of his worship among them, are striking features of the Iliad. 
 
 Coins of Neandreia were found at Assos. The writer can quote, in reference to 
 this subject, no publication more recent than that of Borrell, entitled Unedited Greek 
 Coins, in The Numismatic Chronicle, vol. v. 1. London, 1843. Compare also T. E. 
 Mionnet, Description des Medailles Antiques, vol. v. Paris, 1830. 
 
 '■■^ A coin of this kind is engraved in the work of Sestini which is quoted in the 
 following note, pi. add. iii. The head on the obverse, horned and bearded, is 
 erroneously held by the author to be that of Pan. 
 
 D. Sestini, Descrizione delle medaglie Greche. Parte seconda. Firenze, 1829. 
 On the worship of Apollon in Alexandreia Troas see, also, the inscription from 
 that place, published as No. 3577 of the Corpus Inscr. Grace; together with Boeckh's 
 remarks thereon. 
 
FROM THE SITE OF NEANDREIA. 
 
 35 
 
 Concerning the political history of the place, it is not here neces- 
 sary to enter into detail. Its chief interest attaches to the appearance 
 of Neandreia among the towns tributary to Athens, in the well- 
 known inscription, dating to the third quarter of the fifth century, 
 which has been published by Rhangabe/'^ The amount of the contri- 
 bution, mentioned with each occurrence of the name, permits an inter- 
 esting comparison with that paid by the neighboring towns of Assos, 
 Sigeion, Kebrene, Gargara and I.amponeia. The efforts of this con- 
 federation were not successful, but the second subjugation of the 
 Troad by the Persians was soon followed by the Asiatic conquests of 
 the Greeks, and by the consequent removal of the population of 
 Neandreia to increase that of the thriving port of Alexandreia Troas. 
 
 Thus this remote fortress, deserted by the generation succeeding 
 Alexander the Great, has remained an undisturbed ruin for twenty- 
 two centuries. Romans, Goths, Armenians and Franks claimed the 
 site as their own, the Byzantine Greeks were finally driven altogether 
 from the land by the Seljukian and Ottoman Turks, and the first 
 heed paid to the overthrown monuments of the ancient town should 
 be credited to the Archaeological Institute of a continent not dreamed 
 of by Kalchas or Kassandra. 
 
 Joseph Thacher Clarke. 
 
 Brugsch, in his Troy and Egypt (appendix ix to Schliemann's i/tos, quoted above) 
 suggests that the Trojan Kolonai is to be identified with the Kerena or Kelena 
 mentioned in an inscription engraved upon the walls of a pylon of the temple of 
 Medinet Abou, which gives a list of thirty-nine towns of the Asiatic coast and the 
 neighboring islands whose contingents were defeated by Rameses III in the 13th cen- 
 tury B. c. This theory is rendered particularly attractive by the fact, now placed almost 
 beyond question, that both the Dardanians and the Assians fought against Rameses II 
 only a century before, and are named in the famous poem of Pentaur. Nevertheless, 
 a careful consideration of the context, — especially of the names of those places which 
 are recognizable with some degree of probability, — leads the writer to believe that 
 the town in question must have been situated at least as far south as Kypros. The 
 statement made by Brugsch in his Geschichte Aegyptens unter den Pharaonen (vol. ii. 
 Leipzig, 1878), that the Kerena of the Egyptians was Kerynia, appears much more 
 reasonable. 
 
 '^A. Rhizos Rhankabes, Antiquites Hellenigues. Athenes, 1842-55. Vol. i. No. 
 236, etc.; most recently in the Corpus Inscr. Attic. Vol. i. No. 226.