TROAD AND TROY. The Troad (-r) Tpoias), or land of Geogra-Troy, is the north-western promontory of Asia Minor. The name "Troad" is never used by Homer,—who callstioll*" the land, like the city, Tpoir),—but is already known to Herodotus. The Troad is bounded on the north by the Hellespont and the westernmost part of the Propontis, on the west by the ^Egean Sea, and on the south by the Gulf of Adramyttium. The eastern limit was variously defined by ancient writers. In the widest acceptation, the Troad was identified with the whole of western and south-western Mysia, from the ^Esepus, which flows into the


Propontis a little west of Cyzicus, to the Caicus, which flows into the iEgean south of Atarneus. But the true eastern boundary is undoubtedly the range of Ida, which, starting from near the south-east angle of the Adramyttian Gulf, sends its north-western spurs nearly to the coast of the Propontis, in the region west of the vEsepus~*and east of the Granicus. Taking Ida for the eastern limit, we have the definition which, as Strabo says, best corresponds with the actual usage of the name Troad. Ida is the key to the physical geography of the whole region ; and it is the peculiar character which this mountain-system imparts to the land west of it that constitutes the real distinctness of the Troad from the rest of Mysia. Nature has here provided Asia Minor with an outwork against invaders from the north-west ; and as in the dawn of Greek legend the Troad is the scene of the struggle between Agamem-non and Priam, so it was in the Troad that Alexander won the battle which opened a path for his further advance.
Natural The length of the Troad from north to south—taking a divisions, straight line from the north-west point, Cape Sigeum (Yeni Shehr), to the south-west point, Cape Lectum (Babà-Calessi) —may be roughly given as forty miles. The breadth, from the middle point of the west coast to the main range of Ida, is not much greater. The whole central portion of this area is drained by the Mendere (the ancient Scamander), which rises in Ida and is by far the most important river of the Troad. The basin of the Mendere is divided by hills into two distinct parts, a southern and a northern plain. The southern—anciently called the Samonian plain —is the great central plain of the Troad, and takes its modern name from Bairamitch, the chief Turkish town, which is situated in the eastern part of it near Ida. It is of an elongated form, the extent from north to south being large in proportion to the average width, and is enclosed by hills which, especially towards the south, are low and undulating. From the north end of the plain of Bair-amitch the Mendere winds in large curves through deep gorges in metamorphic rocks, and issues into the northern plain, stretching to the Hellespont. This is the plain of Troy, which has an average length of seven or eight miles from north to south, with a breadth of some two or three from east to west. The hills which enclose it on the south and east are quite low, and towards the east the acclivities are in places so gentle as to leave the limits of the plain somewhat indefinite. Next to the basin of the Mendere, with its two plains, the best marked feature in the river-system of the Troad is the valley of the Touzla, the ancient Satniois. The Touzla rises in the western part of Mount Ida, south of the plain of Bairamitch, from which its valley is divided by hills ; and, after flowing for many miles almost parallel with the south coast of the Troad, from which, at Assus, it is less than a mile distant, it enters the iEgean about ten miles north of Cape Lectum. Three alluvial plains are comprised in its course. The easternmost of these, into which the river issues from rugged mountains of considerable - height, is long and narrow. The next is the broad plain, which is overlooked by the lofty site of Assus, and which was a fertile source of supply to that city. The third is the plain at the embouchure of the river on the west coast. This was anciently called the Halesian (AXrjq-tov) plain, partly from the maritime salt-works at Tragasas, near the town of Hamaxitus, partly also from the hot salt-springs which exist at some distance from the sea, on the north side of the river, where large formations of rock-salt are also found. Maritime salt-works are still in operation at the mouth of the river, and its modern name (Touzla = salt) Coasts, preserves the ancient association. A striking feature of the southern Troad is the high and narrow plateau which runs parallel with the Adramyttian Gulf from east to west, forming a southern barrier to the valley of the Touzla, and walling it off from a thin strip of seaboard. This plateau seems to have been formed by a volcanic upheaval which came late in the Tertiary period, and covered the limestone of the south coast with two successive flows of trachyte. The lofty crag of Assus, washed by the sea, is like a tower standing detached from this line of mountain-wall. The western coast is of a different character. North of the Touzla extends an undulating plain, narrow at first, but gradually widening. Much of it is covered with the valonia oak (Quercus JEgilops), one of the most valuable products of the Troad. Towards the middle of the west coast the adjacent ground becomes higher, with steep acclivities, which sometimes rise into peaks; and north of these, again, the seaboard subsides towards Cape Sigeum into rounded hills, mostly low.
The timber of the Troad is supplied chiefly by the pine- Natural forests on the slopes of Mount Ida. But nearly all the products, plains and hills are more or less well wooded. Besides the valonia oak, the elm, willow, cypress, and tamarisk shrub abound. Lotus, galingale, and reeds are still plentiful, as in Homeric days, about the streams in the Trojan plain. The vine, too, is cultivated, the Turks making from it a kind of syrup and a preserve. In summer and autumn water-melons are among the abundant fruits. Cotton, wheat, and Indian corn are also grown. The Troad is, indeed, a country highly favoured by nature—with its fertile plains and valleys, abundantly and continually irri-gated from Ida, its numerous streams, its fine west sea-board, and the beauty of its scenery. Under a good government, it could not fail to be exceedingly prosperous. Under Turkish rule, the natural advantages of the land suffice to mitigate the poverty of the sparse population, but have scarcely any positive result.
In the Homeric legend, with which the story of the Early Troad begins, the people called the Troes are ruled by a history; king Priam, whose realm includes all that is bounded by ^a™(^l " Lesbos, Phrygia, and the Hellespont" (II., xxiv. 544), i.e., the whole "Troad," with some extension of it, beyond Ida, on the north-west. According to Homer, the Achaeans under Agamemnon utterly and finally destroyed Troy, the capital of Priam, and overthrew his dynasty. But there is an Homeric prophecy that the rule over the Troes shall be continued by vEneas and his descendants. From the " Homeric " Hymn to Aphrodite, as well as from a passage in the 20th book of the Iliad (75-353)—a passage un-doubtedly later than the bulk of the book—it is certain that in the seventh or sixth century B.C. a dynasty claim-ing descent from .-Eneas reigned in the Troad, though the extent of their sway is unknown. The Homeric tale of Troy is a poetic creation, for which the poet is the sole witness. The analogy of the French legends of Charle-magne warrants the supposition that an Achaean prince once held a position like that of Agamemnon. We may suppose that some memorable capture of a town in the Troad had been made by Greek warriors. But we cannot regard the Iliad in any closer or more exact sense as the historical document of a war. The geographical compact-ness of the Troad is itself an argument for the truth of the Homeric statement that it was once united under a strong king. How that kingdom was finally broken up is unknown. Thracian hordes, including the Treres, swept into Asia Minor from the north-west about the beginning of the seventh century B.C., and it is probable that, like the Gauls and Goths of later days, these fierce invaders made havoc in the Troad. The Ionian poet Callinus has recorded, the terror which they caused further south.
A new period in the history of the Troad begins with the foundation of the Greek settlements. The earliest

Greek and most important of these were Jj"olic. Lesbos and settle- Cyme in iEolis seem to have been the chief points from ments. which the first iEolic colonists worked their way into the Troad. Commanding positions on the coast, such as Assus and Sigeum, would naturally be those first occupied; and some of them may have been in the hands of ^Eolians as early as the 10th century B.C. It appears from Hero-dotus (v. 95) that about 620 B.C. Athenians occupied Sigeum, and were resisted by iEolic colonists from Myti-lene in Lesbos, who had already established themselves in that neighbourhood. Struggles of this kind may help to account for the fact noticed by Strabo, that the earlier colonies had often migrated from one site in the Troad to another. Such changes of seat have been, he observes, frequent causes of confusion in the topography; and the fact has an important bearing on attempted identifications of the more obscure ancient sites.
Among the Greek towns in the Troad, three stand out Greek with especial prominence—Ilium in the north, Assus in Ilium, the south, and Alexandria Troas in the west. The site of the Greek Ilium is marked by the low mound of Hissarlik ("place of fortresses") in the Trojan plain, about three miles from the Hellespont. The early Greek settlers in the Troad naturally loved to take Homeric names for their towns. The fact that Homer places the town of Dardania far inland, on the slopes of Ida, did not hinder the founders of the iEolic Dardanus from giving that name to their town on the shores of the Hellespont. The site of the historical Thymbra, again, cannot be reconciled with that of the Homeric Thymbra. Similarly, the choice of the name Ilion in no way justifies the assumption that the Greek settlers found that spot identified by tradition with the site of the town which Homer calls Ilios. It does not even warrant the hypothesis that they found a shrine of Athene Ilias existing there. For them, it would be enough that the sounding name could be safely appro-priated,—the true site of Homeric Ilias being forgotten or disputed,—and that their town was at least in the neigh-bourhood of the Homeric battlefields. The Greek Ilium may have been founded about 700 B.C. It is noticeable that no ancient writer suggests a later date than the time of Crcesus (c. 550 B.C.); and Strabo says that the establishment of the colony at Hissarlik—after previous occupation of a different site—took place "in the time of the Lydians " (eirl AvSwv). It would be reasonable to infer that the Greek Ilium' preserved some well-marked traces of Lydian influence, perhaps in architecture or art, perhaps in manners or traditions. The traces of Lydian workmanship found in the excavations at Hissarlik are thus easily explained, without recourse to the shadowy hypothesis of a distinct Lydian settlement on the spot. When Xerxes visited the Trojan plain, he "went up to the Pergamon of Priam," and afterwards sacrificed to the Ilian Athene (Herod., vii. 42). It is doubtful whether the " Pergamon " meant was at the Greek Ilium, or at another site (to be mentioned presently), Bunarbashi; strong reasons in favour of the latter have lately been adduced by Mr George Nikolaides, in his 'IA«£Sos SxpaT^yiK-r) Atao-Kevi). In the 4th century Ilion is mentioned among the towns of the Troad which yielded to Dercyllidas (399 B.C.), and as captured by Charidemus (359 B.C.). It pos-sessed walls, but was a petty place, of little strength. In 344 B.C. Alexander, on landing in the Troad, visited Ilium. In their temple of Athene the Ilians showed him arms which had served in the Trojan war, including the shield of Achilles. Either then, or after the battle of Granicus, Alexander directed that the town should be enlarged, and should have the rank of " city," with political independence, and exemption from tribute. The battle of Ipsus (301 B.C.) added north-western Asia Minor to the dominions of Lysimachus, who executed the intentions of Alexander. He gave Ilium a wall 5 miles in circum-ference, incorporating with it some decayed towns of the neighbourhood, and built a handsome temple of Athene. In the 3d century B.C. Ilium was the head of a federal league (KOIVOV) of free Greek towns, which probably in-cluded the district from Lampsacus on the Hellespont to Gargara on the Adramyttian Gulf. Twice in that century Ilium was visited by Gauls. On the first occasion (278 B.C.) the Gauls, under Lutarius, sought to establish a stronghold at Ilium, but speedily abandoned it as being too weak for their purpose. Forty years later (218 B.C.) Gauls were brought over by Attalus I. to help him in his war against Achseus. After deserting his standard they proceeded to pillage the towns on the Hellespont, and finally besieged Ilium, from which, however, they were driven off by the troops of Alexandria Troas. At the beginning of the 2d century B.C. Ilium was in a state of decay. As Demetrius of Scepsis tells us, the houses " had not even roofs of tiles," but merely of thatch. Such a loss of prosperity is sufficiently explained by the incursions of the Gauls and the insecure state of the Troad during the latter part of the 3d century. The temple of the Ilian Athene, however, retained its prestige. In 192 B.C. Antiochus the Great visited it before sailing to the aid of the iEtolians. In 190 B.C., shortly before the battle of Magnesia, the Romans came into the Troad. At the moment when a Roman army was entering Asia, it was politic to recall the legend of Roman descent from ./Eneas. Lucius Scipio and the Ilians were alike eager to do so. He offered sacrifice to the Ilian Athene; and after the peace with Antiochus (189 B.C.) the Romans annexed Rhoeteum and Gergis to Ilium, " not so much in reward of recent services, as in memory of the source from which their nation sprang." The later history of Ilium is little more than that of Roman benefits. A disaster befell the place in 85 B.C., when Fimbria took it, and left it in ruins; but Sulla presently caused it to be rebuilt. Augustus, while confirming its ancient privileges, gave it new terri-tory. Caracalla (211-217 A.D.) visited Ilium, and like Alexander paid honours to the tomb of Achilles. The latest coins found on the site are those of Constantius II. (337-361). In the 4th century, as some rhetorical " Letters " of that age show, the Ilians still did a profit-able trade in attracting tourists by their pseudo-Trojan memorials. After the 4th century the place is lost to view. But we find from Constantine Porphyrogenitus (911-959) that in his day it was one of the places in the Troad which gave names to bishoprics.
While the Greek Ilium at Hissarlik owed its importance Assus. to a sham pretension, which amused sight-seers and occa-sionally served politicians, Assus, on the south coast, has an interest of a more genuine kind, and is, indeed, a better type of ancient town-life in the Troad. Its situation is one of the most magnificent in all the Greek lands. The seaward faces of the isolated and sea-washed rock on which Assus stood are carved to south and south-west into terraces. The natural cleavage of the trachyte into joint planes had already scarped out shelves which it was comparatively easy for human labour to shape; and so, high up on this cone of trachyte, the Greek town of Assus was built, with its colonnades, baths, theatre, its public walks and its monuments of the dead, mounting tier above tier, till the summit of the crag was crowned with a Doric temple of Athene. The view from the summit is not only very beautiful but also of great historical interest. In front is Lesbos, one of whose towns, Methymna, is said to have sent forth the founders of Assus, as early, perhaps, as 1000 or 900 B.C. The whole south coast-line of the Troad is seen, and in the south-east the ancient territory

of Pergamum, from whose masters the possession of Assus passed to Rome by the bequest of Attalus III. (133 B.C.). The great heights of Ida rise in the east. Northward the Touzla is seen winding through its rich valley from a rocky defile in the east to the oak-forests in the western hills. This valley was traversed by the road which St Paul must have followed when he came overland from Alexandria Troas to Assus, leaving his fellow-travellers to proceed by sea. The north-west gateway of Assus, to which this road led, is still flanked by two massive towers, of Hellenic work, and of an age which leaves no doubt that they are the same between which St Paul entered the town. On the shore below, the ancient mole at which he embarked for Mytilene with his companions can still be traced by large blocks under the clear water. Assus affords the only harbour on the 50 miles of coast between Cape Lectum and the east end of the Adramyttian Gulf ; hence it must always have been the chief shipping-place for the exports of the southern Troad. Too much off the highways to become a centre of import trade, it was thus destined to be a commercial town, content with a modest provincial prosperity. The great natural strength of the site protected it against petty assailants; but, like other towns in that region, it has known many masters,— Lydians, Persians, the kings of Pergamum, Romans, and Ottoman Turks. From the Persian wars to about 350 B.C. Assus enjoyed at least partial independence. It was about 348-345 B.C. that Aristotle spent three years at Assus with Hermeas, an ex-slave who had succeeded his former master Eubulus as despot of Assus and Atarneus. Aristotle has left some verses from an invocation to Arete (Virtue), commemorating the worth of Hermeas, who had been seized by Persian treachery and put to death. Under its Turkish name of Beihram, Assus is still the commercial port of the southern Troad, being the place to which loads of valonia (acorn-cups for tanning) are con-veyed by camels from all parts of the country. The recent excavations at Assus, conducted by explorers repre-senting the Archaeological Institute of America, have yielded results far more valuable for the history of Greek art and architecture than any excavations yet undertaken in the Troad. The sculptures form one of the most important links yet discovered between Oriental and early Greek art, especially in respect of the types of animals. The later Hellenic town-walls of Assus also well repay the new study which they have received. With their ramparts, towers, and posterns they form the finest and most instructive extant specimen of Greek military engineering. The director of the exploration, Mr J. T. Clarke, published in 1882 an excellent report on the work so far as it had then been carried. Alex- Alexandria Troas stood on the west coast at nearly its middle point, a little south of Tenedos. It was built by Antigonus, perhaps about 310 B.C., and was called by him Antigonia Troas. Early in the next century the name was changed by Lysimachus to Alexandria Troas, in honour of Alexander's memory. As the chief port of north-west Asia Minor, the place prospered greatly in Roman times, and the existing remains sufficiently attest its former importance. The site is now covered with valonia oaks; but the circuit of the old walls can be traced, and in several places they are fairly well preserved. They had a circumference of about 6 English miles, and were fortified with towers at regular intervals. Remains of some ancient buildings, including a bath and gymnasium, can be traced within this area. The harbour had two large basins, now almost choked with sand. A Roman colony was sent to the place, as Strabo mentions, in the reign of Augustus. The abridged name "Troas" (Acts xvi. 8) was probably the current one in later Roman times. The site is now called Eski Stambul.
Many classical sites of less note in the Troad have been Other identified with more or less certainty. Neandria seems to ancient be rightly fixed by Mr F. Calvert at Mount Chigri, a hillsites-not far from Alexandria Troas, remarkable for the fine view of the whole Troad which it commands. Cebrene has been conjecturally placed in the eastern part of the plain of Bairamitch, Palsescepsis being further east on the slopes of Ida, while the new Scepsis was near the site of Bairamitch itself. The evidence for this, however, is ambiguous. At the village of Kulaklee, a little south of the mouth of the Touzla, some Corinthian columns and other fragments mark the temple of Apollo Smintheus and (approximately) the site of the Homeric Chryse. Colonse was also on the west coast, opposite Tenedos. Scamandria occupied the site of Eneh, in the middle of the plain of Bairamitch, and Cenchreae was probably some distance north of it. The shrine of Palamedes, mentioned by ancient writers as existing at a town called Polymedium, has been discovered by Mr J. T. Clarke on a site hitherto unvisited by any modern traveller, between Assus and Cape Lectum. It proves to have been a sacred enclosure (témenos) on the acropolis of the town; the statue of Palamedes stood on a rock at the middle of its southern edge. Another interesting discovery has been made by Mr Clarke,—viz., the existence of very ancient town walls on Gargarus, the highest peak of Ida.
The modern discussion as to the site of Homeric Troy Site of may be considered as dating from Lechevalier's visits to Homeric the Troad in 1785-86. Homer describes Troy as "a greatTroy' town," "with broad streets," and with a high acropolis, or "Pergamus," rising above it, from which precipitous rocks descend abruptly to the plain beneath. These are the precipices over which the Trojans proposed to hurl the wooden horse, "when they had dragged it to the summit." Homer marks the character of the acropolis by the epithets "lofty," "windy," and more forcibly still by "beetling." One site in the Trojan plain, and one only, satisfies this most essential condition. It is the hill at its southern edge called the Bali Dagh, above the village of Bunár-bashi. It has a height of about 400 feet, with sheer precipices descending on the south and south-west to the valley of the Scamander (Mendere). Remains found upon it—though it has never yet been thoroughly explored— show it to have been the site of an ancient city. Homer describes two natural springs as rising a little to the north-west of Homeric Troy. A little to the north-west of Bunárbashi these springs still exist. ." This pair of rivu-lets are the immutable mark of nature by which the height towering above is recognized as the citadel of Ilium " (E. Curtius).
The low mound of Hissarlik—the site of the Greek Eejec-Ilium—stands only 112 feet above the level of the opentionof plain in which it is situated. To call it " beetling" (¿fipvóecro-a) would have been a travesty of poetical licence on which no poet could have ventured, and to describe it as " lofty " or " windy " would have been not less strange. There are no natural springs near it, such as Homer mentions. The iEolic settlers, having called the place Ilion, naturally persisted in maintaining its identity with Troy. Polemon, a native of the Greek Ilium, who lived about 200 B.C., declared that his fellow-townsmen could show the very stone on which Palamedes

had given lessons in the game of draughts. The only
other ancient writer who is known to have admitted the
Ilian claim is Hellanicus of Lesbos (c. 482-397 B.C.), who,
as Strabo remarks, wished " to gratify the Ilians, as is his
wont." Like the Ilians, Hellanicus was of iEolian origin;
and in compiling the local legends of various places he is
known to have been wholly uncritical, merely repeating
By an- what was told to him as he had heard it. On the other
cient _ hand, the claim of the Greek Ilium to stand on the site of
criticism; rpr0y wag ,jecisiyepy rejected by the general consent of
those ancient writers who had any claim to critical
authority. The orator Lycurgus (c. 332 B.C.) speaks of
the site of Troy as desolate, and this at a moment when
the recent visit of Alexander the Great to the Greek Ilium
(334 B.C.) had drawn attention to the claim made by its
inhabitants. Demetrius, a native of Scepsis in the Troad,
who flourished about 160 B.C., wrote, a book entitled
Tpon/cos AtaKooytos ("The Marshalling of the Trojans"),
an exhaustive commentary on the catalogue of the Trojan
forces in the second book of the Iliad. Demetrius knew
the topography of the Troad as thoroughly as he knew the
text of Homer. The extant notices of his work, which
had a great reputation in antiquity, warrant the belief
that he was not only learned but acute. In the Diacosmus,
which was the chief work of his life, he must have
bestowed much thought on the question as to the site of
Homeric Troy,—the central point of his subject. He
pronounced decidedly, as we know from Strabo, against
the claim of the Greek Ilium. It has been suggested
that Demetrius rejected the Ilian claim because, as a
native of Scepsis, he was jealous of Ilium,—a suggestion
which is not only absurd in itself, since it assumes that
such a motive would have induced Demetrius to mar his
life's work, but also betrays ignorance of Strabo's text.
Scepsis was not a possible claimant of the contested
honour, since it was not in the plain of Troy but in the
plain of Bairamitch; and further, Demetrius had already
provided in another manner for the Homeric dignity of
. Scepsis by making it the royal seat of iEneas on the
strength of its position relatively to Lyrnessus. The
verdict of Demetrius against the Ilian claim was also the
general verdict of the other ancient writers consulted by
Strabo, as the latter's language shows. From the passage
in which Strabo notices the various definitions of the Troad
(xiii. § 4) it appears that among such writers were the
following historians and geographers :—Charon of Lamp-
sacus (nor. 500 B.C.), Damastes of Sigeum (400 B.C.),
Scylax of Caryanda (350 B.C.), Ephorus of Cyme (340
B.C.), Eudoxus of Cyzicus (130 B.C.). It is to such writers
as these that Strabo refers when he indicates the general
consent of his authorities. In favour of the claim of the
Greek Ilium, on the other hand, there are only two literary
witnesses, and these, as we have seen, are alike worthless.
Equally valueless from a critical point of view is the fact
that the Ilian claim was sometimes allowed by soldiers
or statesmen who wished to utilize Trojan memories.
They required an official Troy, and they cared not where
they found it. Nothing could more curiously illustrate
the extreme poverty of the case for the Greek Ilium than
the fact that some of its advocates have been reduced to
arguing as if Alexander and Lucius Scipio, when they
led their armies through the Troad, had been conducting
archaeological excursions, and as if their acquiescence in a
convenient local myth had the weight of independent
critical testimonies.
By In negativing the Ilian claim the conclusion of ancient
modern, criticism has been confirmed by a great preponderance of modern opinion. Since Lechevalier visited the Troad in 1785-86 an overwhelming majority of competent judges have favoured his belief that the Bali Dagh above Bunarbashi was the Pergamus of the Homeric poet's conception. Before Leake's visit this opinion had been expressed by Choiseul-Gouffier, Morritt, Hawkins, Gell, and Hamilton. Leake spoke with a decision which derives additional weight from the habitual sobriety of his acute judgment, and from the care with which, in this case, he had ex-amined the alleged objections to the view which he finally adopted. He remarks that no one accustomed to observe the sites of ancient Greek towns could fail to fix on Bunar-bashi " for the site of. the chief place of the surrounding country." So Mr Tozer, in his Highlands of Turkey, says: "A person accustomed to observe the situation of Hellenic cities would at once fix on this as far more likely to have recommended itself to the old inhabitants of the country than any other in the neighbourhood." Count von Moltke has expressed the same opinion, that "he knew no other site in the Trojan plain for a chief town of ancient time." Another supporter of Bunarbashi is Forchhammer. Another is Kiepert. The opinion of Ernst Curtius has been already cited. But space precludes more names; it is enough to say that the correspondence of the Bali Dagh with the Homeric Pergamus—a correspondence absolutely unique in the Trojan plain—has been recognized with virtual una-nimity by modern travellers who have patiently inspected the scenery of the Iliad, having competent knowledge, and being free from bias in favour of a theory formed before their visit. Partial excavations on the summit of the Bali Dagh have been more than once undertaken, with the result of discovering ancient walls. Pottery, too, has been found there, part of which is allowed on all hands to be probably as old, at least, as 900 B.C. But the Bali Dagh has never yet been explored with any approach to thoroughness.
The result of the excavations conducted by Dr Schlie- The re-mann on the mound of Hissarlik has been to lay bare mains at the remains of the Greek Ilium, and also, below these, Hissarlik some prehistoric remains of a rude and poor kind. In Troy, his first book on the subject, the explorer held that the remains of the Greek Ilium ceased at a depth of 6 feet below the surface, and that all the" other remains, down to 52J feet, were prehistoric. He distinguished the latter into five groups, representing five prehistoric "cities" which had succeeded each other on the site; and in his second work, Ilios, he added to these a sixth prehistoric city, on the strength of some scanty vestiges of supposed Lydian workmanship, found at a depth of 6J feet. In both books, Homeric Troy was identified with the third prehistoric city from the bottom, which was supposed to have been destroyed, though not totally, by fire. Professor Jebb was the first to show (1) that the lines of demarcation between the alleged prehistoric strata, as drawn in Ilios, could not be accurate, and (2) that, if any part of the pre-historic remains could be supposed to represent Homeric Troy, it must be that part which Dr Schliemann had called the second city from the bottom, and the destruction of which by fire appeared to have been total. In 1882 the architects employed by Dr Schliemann proved that the stratification given in Ilios had in fact been incorrect. The errors, too, affected precisely that region of the deposit which was most important to the Trojan hypothesis, viz., the lower strata. In Dr Schliemann's third volume, Troja, these errors were admitted; and Troy was now identified, no longer with the third city, but with the second, of which the supposed area was now enlarged. Another fact to which the English critic had drawn attention was that the remains of the Greek Ilium must extend to a considerably greater depth than 6 feet below the surface. Further examination confirmed this view also. It showed that the remains on .the mound at Hissarlik belong to the following periods or groups. (1) At the top, the remains of the Greek Ilium as it existed in the Boman age, i.e., as rebuilt

after its destruction by Fimbria in 85 B.C. (2) A city which, like the former, extended beyond the mound of Hissarlik (its acropolis) over .the adjacent plain. This corresponds with the Greek Ilium of the Macedonian age, as embellished and enlarged by Lysimachus, c. 300 B.C. (3) A smaller city, probably confined to the mound. Here we may recognize the Greek Hium as it existed before the Macedonian age. It was a small and poor place, as appears from the known incidents of its history in the 5th and 4th centuries B.C., owing its chief importance to the shrine of Athene Ilias. (4) A petty town or village, confined to the mound, and poorly built. The evidence of architecture fails to decide whether it was Hellenic or not; if Hellenic, it might represent the primitive settlement of the zEolic colonists, perhaps c. 700 B.C. It was a small house in this village that Dr Schliemann at first identified with Priam's palace. The ground-plan shows four rooms, of which the largest measured 24 feet 4 inches by 12 feet. (5) A large town, to which the mound was only acropolis, and which extended to some distance south and south-east over the plain. These remains are unquestionably prehistoric. (6) A few remains of a small settlement which, if indeed distinct from No. 5, preceded it. The reason for distinguishing 6 from 5 is that some of the acropolis buildings of 5 are above those of 6, and seem to have been built on carefully levelled ground. Apart from architectural evidence, objects found in the excavations prove that the remains of the historical age extend much below 6 feet. One of these was a terra-cotta disk, stamped with the head of a warrior, in an advanced style of workmanship, found at 26 feet 3 inches below the surface (Troy, p. 294). Another is a terra-cotta ball, found at 26 feet, which cannot be older than c. 360 B.C. Then, at 20 feet, was found another terra-cotta, marked with the Greek letter P. A piece of ivory, belong-ing to a seven-stringed lyre, and therefore not older than c. 660 B.C., was found at 26 feet. Thus we have at Hissarlik the remains of the Greek Ilium in three successive phases,—Roman, Macedonian, and iEolic, and below these the remains of at least one prehistoric settlement, the age and origin of which are unknown. Their re- We can no longer either prove or disprove that these lation to prehistoric remains are those of a town which was once Homer, taken after a siege, and which originally gave rise to the legend of Troy. But most certainly it is not the " lofty " Troy of which the Homeric poet was thinking when he embodied the legend in the Iliad. The conception of Troy which dominates the Iliad is based on the site at Bunarbashi, and suits no other. The sole phrase in the epic which favours Hissarlik occurs in book xx. (216 sq.), where Dardania is said to have been built on the spurs of Ida, when Ilios " had not yet been built in the plain "; and this phrase occurs in a passage which, as the best recent critics agree, is one of the latest interpolations in the Iliad, having been composed after the Greek Ilium had actually arisen "in the plain." Its purpose was the same as that which appears in the Hymn to Aphrodite, viz., to glorify reputed descendants of iEneas, and it probably belongs to the same age, the 7th century B.C. The tactical data of the Iliad—those derived from the incidents of the war—cannot be treated with such rigour as if the poem were a military history. But Nikolaides has shown that~they can at least be brought into general agreement with the site at Bunarbashi, while they are hopelessly incompatible with Homeric Hissarlik. The Iliad makes it clear that the general concep- description of the Trojan plain was founded on accurate of knowledge. At this day all the essential Homeric features can be recognized. And it is probable that the poet who created the Troy of the Iliad knew, personally or by description, a strong town on the Bali Dagh above Bunar-bashi. The legend of the siege may or may not have arisen from an older town at Hissarlik, which had then disappeared. The poet might naturally place his Troy in a position like that of the existing strong city on the Bali Dagh, giving it a "beetling" acropolis and handsome buildings, while he also reproduced the general course of the rivers and that striking feature,—an indelible mark of the locality,—the natural springs at the foot of the hill, just beyond the city gates on the north-west. But, while he thus imagined his Troy in the general likeness of the town on the Bali Dagh, he would retain the privilege of a poet who was adorning an ancient legend, and whose theme was a city that had long ago vanished. Instead of feeling bound to observe a rigorous accuracy of local detail, he would rather feel impelled to avoid it ; he would use his liberty to introduce some traits borrowed from other scenes known to him, or even from imagination. To this extent, and in this sense, his topography would be eclectic. Such a consideration might suffice to explain the fact, well known to those who have studied this question on the spot, that neither Bunarbashi nor any other one site can be harmonized with every detail of the poem. The recommendations of Bunarbashi are, first, that it satisfies the capital and essential conditions, while no other site does so, and secondly, that the particular difficulties which it leaves unsolved are relatively slight and few. This character of Homeric topography becomes still easier to understand, if, as most critics would now concede, our Iliad contains work of various hands and ages. Few questions, perhaps, of equal literary interest have been so much confused by inattention to the first conditions of the problem. The tale of Troy, as the Iliad gives it, is essentially a poetical creation ; and we have no evidence other than the Iliad. That is, our sole data are (1) of the mythical class, (2) of inadequate precision, and (3) of un-certain origin. But they show a general knowledge of the ground ; and.the question is how far particular features of the ground can be recognized in the poem. It may be doubted whether the case admits of any solution more definite than that which has been indicated above.
Bibliography.—1. "Works dealing with the Troad generally,— Strabo, bk. xiii. ch. 1, is the principal source for the ancient Troad. Of books by modern travellers in Asia Minor the following may be-mentioned :—Philip Barker Webb, in the Italian Biblioteca Acerbi, June and July, 1821, whose studies are better known through the French edition, Topographie de la Troade, 1844 ; W. M. Leake, Journal of a Tour in Asia Minor, London, 1S24 ; P. de Tchihatchef,. Asie Mineure, &c, Paris, 1853-69; R. Virchow, "Beiträge zur' Landeskunde der Troas," in Trans, of Berlin Acad., 1879 ; H. F. Tozer, The Highlands of Turkey, 1869 ; H. Schliemann, Meise der Troas in Mai, 1881 ; Joseph T. Çlarke, Report on the Investigations at Assos, Boston, U.S.A., and London, 1882, including "Notes on the Geology and Topography of the Troad " by J. S. Diller, and on "Bunarbashi," &c, by W. C. Lawton and C. H. Walker. J. T. Clarke's "Notes on Greek Shores,"in the Report of the Archaeo-logical Institute of America for 1880, are also valuable.
2. For the question as to the- site of Troy, see—Lechevalier,
Voyage de la Troade, Paris, 1802 ; Gustave D'Eichthal, Le Site de
Troie selon Lechevalier, &c., Paris, 1875 ; H. Schlieniann's Troy
(1875), Ilios (1880), Troja (1884), which contain many good plans
and illustrations ; E. Brentano, Zur Lösung der troianischen Frage,
Heilbronn, 1881, and Troia und Neu-Ilion, ibid., 1882 ; R. C. Jebb,
" Schliemann's Ilios," itiEdinb. Rev., No. eeexiv., April, 1881 ; Id.,
"Homeric and Hellenic Ilium," in Journ. of Hellenic Shidies, vol.
ii. pp. 7-43,1881 ; Id., " The Ruins at Hissarlik," ibid., iii. 185-217,
1882 ; Id., "Homeric Troy," in Fortnightly Review, April, 1884 ;
G. Nikolaides, 'IAidôos ^TparriyiKri Aiao-Kevi), Athens, 1883 ; P. W.
Forchhammer, Erklärung der Ilias, auf Grund der in der beigege-
benen Original-Karte von Spratt and Forchhammer dargestellten-
topischen und physischen Eigenthümliclikeiten der Troischen Ebene,
Kiel, 1884; and W. J. Stillman, "Les Découvertes de Schliemann,"
in the Journal L'Homme, Paris, October, 1884. (R. C. J.)
LEGEND OF TROY.
According to Greek legend, the oldest town in the-Troad was that founded by Teucer, who was a son of the river Scamander and the nymph Idtea. Tzetzes says

that the Scamander in question was the Scamander in Crete, and that Teucer was told by an oracle to settle wherever the " earth-born ones " attacked him. So when he and his company were attacked in the Troad by mice, which gnawed their bow-strings and the handles of their shields, he settled on the spot, thinking that the oracle was fulfilled. He called the town Sminthium and built a temple to Apollo Sminthius, the Cretan word for a mouse being sminthius. In his reign Dardanus, son of Zeus and the nymph Electra, daughter of Atlas, in con-sequence of a deluge, drifted from the island of Samo-thrace on a raft or a skin bag to the coast of the Troad, where, having received a portion of land from Teucer and married his daughter Batea, he founded the city of Dardania or Dardanus on high ground at the foot of Mount Ida. On the death of Teucer, Dardanus succeeded to the kingdom and called the whole land Dardania after himself. He. begat Erichthonius, who begat a son Tros by Astyoche, daughter of Simois. On succeeding to the throne, Tros called the country Troy and the people Trojans. By Callirrhoe, daughter of Scamander, he had three sons,—Uus, Assaracus, and Ganymede. From Ilus and Assaracus sprang two separate lines of the royal house,—the one being Ilus, Laomedon, Priam, Hector; the other Assaracus, Capys, Anchises, ^Eneas. Ilus went to Phrygia, where, being victorious in wrestling, he received as a prize from the king of Phrygia a spotted cow, with an injunction to follow her and found a city wherever she lay down. The cow lay down on the hill of the Phrygian Ate; and here accordingly Ilus founded the city of Ilios. It is stated that Dardania, Troy, and Ilios became one city. Desiring a sign at the foundation of Ilios, Ilus prayed to Zeus and as an answer he found lying before his tent the Palladium, a wooden statue of Pallas, three cubits high, with her feet joined, a spear in her right hand, and a distaff and spindle in her left. Ilus built a temple for the image and wor-shipped it. By Eurydice, daughter of Adrastus, he had a son Laomedon. Laomedon married Strymo, daughter of Scamander, or Placia, daughter of Atreus or of Leucippus. It was in his reign that Poseidon and Apollo, or Poseidon alone, built the walls of Troy. In his reign also Hercules besieged and took the city, slaying Laomedon and his children, except one daughter Hesione and one son Pod-arces. The life of Podarces was granted at the request of Hesione; but Hercules stipulated that Podarces must first be a slave and then be redeemed by Hesione; she gave her veil for him; hence his name of Priam (from priasthai to buy). Priam married first Arisbe and after-wards Hecuba and had fifty sons and twelve daughters. Among the sons were Hector and Paris, and among the daughters Polyxena and Cassandra. To recover Helen, whom Paris carried off from Sparta, the Greeks under Agamemnon besieged Troy for ten years. (See ACHILLES, AGAMEMNON, AJAX, HECTOR, HELEN, PARIS.) At last they contrived a wooden horse, in whose hollow belly many of the Greek heroes hid themselves. Their army and fleet then withdrew to Tenedos, feigning to have raised the siege. The Trojans conveyed the wooden horse into Troy; in the night the Greeks stole out, opened the gates to their returning friends, and Troy wras taken.
See Homer, II., vii. 452 sq., xx. 215 sq., xxi. 446 sq. ; Apollo-dorus, ii. 6, 4, iii. 12 ; Diodoms, iv. 75, v. 48 ; Tzetzes, Schol. on Lycophron, 29, 72, 1302 ; Conon, Narrat., 21; Dionysius Halicarn., Antiq. Bom., i. 68 sq. The Iliad deals with a period of fifty-one days in the tenth year of the war. For the wooden horse, see Homer, Od., iv. 271 sq.; Virgil, sEn., ii. 13 sq.
TROGLODYTES (rpojyAoSuTou), a Greek word mean-ing " cave-dwellers." Caves have been widely used as human habitations both in prehistoric and in historic times (see CAVE), and ancient writers speak of Troglodytes in various parts of the world, as in Mcesia near the lower
Danube (Strabo, vii. 5, p. 318), in the Caucasus (Id., xi. 5, p. 506), but especially in various parts of Africa from Libya (Id., xvii. 3, p. 828) to the Red Sea. Herodotus (iv. 183) tells of a race of Troglodyte Ethiopians in inner Africa, very swift of foot, living on lizards and creeping things, and with a speech like the screech of an owl. The Garamantes hunted them for slaves. It has been supposed that these Troglodytes may be Tibbus, who still in part are cave-dwellers. Aristotle also (Hist. An., vii. 12) speaks of a dwarfish race of Troglodytes on the upper course of the Nile, who possessed horses and were in his opinion the Pyg-mies of fable. But the best known of these African cave-dwellers were the inhabitants of the "Troglodyte country" on the coast of the Red Sea, who reached as far north as the Greek port of Berenice, and of whose strange and sav-age customs an interesting account has been preserved by Diodorus and Photius from Agatharchides. They w-ere a pastoral people, living entirely on the flesh of their herds, or, in the season of fresh pasture, on mingled milk and blood. But they killed only old or sick cattle (as indeed they killed old men who could no longer follow the flock), and the butchers were called " unclean "; nay, they gave the name of parent to no man, but only to the cattle of which they had their subsistence. This last point seems to be a confused indication of totemism. They went almost naked; the women wore necklaces of shells as amulets. Marriage was unknown, except among the chiefs, —a fact which agrees with the prevalence of female kin-ship in these regions in much later times. They practised circumcision or a mutilation of a more serious kind. The whole account, much of which must be here passed by, is one of the most curious pictures of savage life in ancient literature.
The Biblical Horim, who inhabited Mount Seir before the Edom-ites, bore a name which means cave-dwellers, and may probably have been a kindred people to the Troglodytes on the other side of the Red Sea. Jerome, on Obadiah 5, speaks of this region as containing many cave-dwellings, and such habitations are still sometimes used on the borders of the Syro-Arabian desert.


Footnotes

The name Assus probably means " dwelling," " town," being con-nected with the Sanskrit vas, "to dwell," which appears in the Greek astu, and also in the ending of such names as Mylasa and Larissa, where in Greek the s is alternatively single or double—an ending which, as Fligier has shown, is found in old town names from India to Dacia. Homer supplies an example in his " steep Pedasus " on the Satniois, and it has been suggested by Mr J. T. Clarke that Pedasus may have been identical in site with the later Assus.


See also Artemidorus in Strabo, xvi. 17, p. 785 sq.


Trogonem (the oblique case) occurs in Pliny (H. N., x. 16) as the name of a bird of which he knew nothing, save that it was mentioned by Hylas, an augur, whose work is lost; but some would read Trygoncm (Turtle-Dove). In 1752 Mohring (Av. Genera, p. 85) applied the name to the "Curucui" (pronounced "Suruqua" fide Bates, Nat. Amazons, i. p. 254) of Marcgrave (Hist. Nat. Brasiliss, p. 211), who described and figured it in 1648 recognizably. In 1760 Brisson (Ornithologie, iv. p. 164) adopted Trogon as a generic term, and, Linnaeus having followed his example, it has since been universally accepted.