a large part of northern Greece and the western islands. Thucydides was right in conceiving him as much more than the temporary leader of a confederacy formed for the special purpose of the war. He was over-lord of Peleus, whose kingdom embraced Phthia and Hellas, and of Odysseus, who ruled over the Ionian islands with the exception of Corfù. Mr Leaf has shown with admirable lucidity how overwhelming are the arguments in favour of Dr Dörpfeld's identification of Homeric Ithaca with Leucas (Santa Maura). The four islands under the sway of Odysseus were Zacynthus, Dulichion, Same, and Ithaca. Zacynthus preserved its name; Dulichion is Cephallenia; Same is Thiaki, the Ithaca of historical Greece; Ithaca is Leucas. These identifications render the geography of the Odyssey completely intelligible and coherent; on the old theory we are involved in a series of insuperable difficulties. Our attitude to tradition is largely a matter of temperament; and there will probably always be some who will prefer to impose on the poet any number of inconsistencies and incongruities rather than sacrifice the tradition of the identity of the Homeric with the later Ithaca. It has been conjectured that in postHomeric days northern invaders seized Leucas, when the Ithacans were driven across to Same and carried their name with them. This may be the explanation. Calabria, which was once the name of the heel of Italy, is now the name of the toe. We know when the change happened, in the seventh century A.D., and why. If the early history of the Middle Ages were as blank to us as the dark period of Greece, we should find it far more puzzling to account for the migration of the Calabrian name than it is to discover a probable reason for the new nomenclature of the Ionian islands.

Mr Leaf has established on a firm basis the value of Homer, within certain wide limits, as a historical source. He has shown that the geographical and topographical details of the Homeric picture conform to fact so far as they can be controlled, and otherwise are self-consistent. Archæological discoveries have proved that in their picture of civilisation the poems are also true to fact, apart from some inevitable anachronisms. Such truth could not have been achieved if a poet of the tenth

century had constructed his epic from pieces of floating tradition. The existence of a body of earlier minstrelsy, coming down from the Achæan period itself, is the only hypothesis which will reasonably explain the data. Moreover the Trojan War must now take its place as an indubitable historical event; and it must have been waged in much the same way as Homer conceived it, 'by a process of wearing down.' The grip of Homeric tradition on reality comes out in the rôle of Lemnos. This island is the natural basis of an army from overseas acting at the mouth of the Hellespont, whether in 1185 B.C. or in 1915 A.D.; and as such it appears in Homer. But, beyond the general setting and conditions, can we hope to discover in Homer anything of the nature of a true story? An epic poem is not a chronicle. In a work in which Clio and Calliope have collaborated, it may seem a hopeless problem to discriminate their contributions. And a full solution would have to determine both the poetical liberties with fact taken by the old minstrels, and the later inventions wrought by the art of Homer into the woof which they bequeathed to him. To penetrate so far into the secrets of the epic is perhaps beyond human powers of analysis, but it may not be impossible to disengage some leading facts.

Probability must indeed be our guide. But arguments of probability which consider what is likely to have happened may profitably be distinguished from those which consider what is likely to have been handed down. It may be argued, for instance, that a war lasting ten years is a poetical fiction, because it appears highly improbable that Agamemnon could have kept together his various contingents far from their homes for so many summers and winters. But this consideration is far from being decisive; there may have been factors in the situation of which we are totally ignorant. The argument for accepting the tradition that an immediate cause of the war was the abduction of an Achæan queen by a Trojan prince is also based on probability, but it is of a different order. The episode is obviously a fundamental part of the original story, as told by the Achæan poets; and it is almost inconceivable that these bards, singing at the courts of the sons and grandsons of the heroes of the war, should have made the whole tale hang on an

incident which was a pure fiction. In the first case the argument for rejecting the tradition depends on our view of what was likely to happen in circumstances with which we are only partly acquainted. In the second case the argument is that the simplest and most satisfactory explanation of the tradition is that it is based on truth. It may be pointed out that to accept the rape of Helen as immediately leading to the expedition against Troy is not inconsistent with the conclusion that the deeper causes of the war were economic; we may compare it with the affair of Epidamnos or with the murder of Serajevo.

In applying the test of probability we have one objective guide, the analogy of other epics which grew up under similar conditions, and for which we have some data to control their treatment of historical facts. Useful hints might be gained from a comparison of the cycle of Merovingian poetry, not preserved indeed in its original form, but copiously used by Fredegarius and the author of the Gesta Francorum.' Mr Leaf has made

frequent use of Prof. Chadwick's instructive work, 'The Heroic Age,' in which the Greek and Teutonic epics are studied and compared. The results of comparisons of this kind point irresistibly to the conclusion that the leading Achæan heroes were not creatures of fiction but men of flesh and blood. The court minstrels of the twelfth century sang of their deeds to descendants who had not forgotten their names. Mr Leaf is assuredly right in asserting the reality of Agamemnon and Menelaus, Nestor and Achilles, Diomede and the Ajaxes. Nor can we logically hesitate to accept as historical the names of their fathers Atreus, Peleus, and the rest. At the point where the poet's knowledge of their ancestry ended, he often introduced a god or an eponymous hero. But, if we go so far, we must go farther. The argument which Mr Leaf applies to the Achæans must also be applied to the principal Trojan heroes. Priam and his father Laomedon, Hector and Paris, Æneas chief of the Dardanians, must have been real people.

The Troes are commonly stated to have been a branch of the Phrygians. It may be doubted whether this is true. In Homer there is no trace of a closer relationship of the Trojans with the Phrygians than with the Lycians; Vol. 226.-No. 448.

C

and it is well to remember that, in the Homeric Hymn to Aphrodite, the Phrygian language is assumed to be unintelligible to a Trojan. It is with the Dardanians that the Trojans seem to have a nearer affinity; and, as the Dardanians must be derived from the central regions of the Balkan Peninsula which retained in later history the Dardanian name, there might seem to be more reason for connecting the Troes with the Illyrians than with the Thraco-Phrygians. However this may be, the Troes were doubtless early immigrants from the Balkan Peninsula. How comes it that their rulers have Greek names? The name of Priam himself is not indeed obviously Greek, but, with its Æolic form Пéppaμoç, it may well be so; and Priam's father was Laomedon. 'Hector' is as Greek as 'Nestor,' and was in later times the name of a prince of Chios. Paris has the second name of Alexandros; and the natural assumption is that 'Paris' was a Phrygian name given to him by his Phrygian mother, Hecuba. The names of other children of Priam who come into the story-Cassandra, Helenus, Deiphobus, etc.-are Greek. We have to choose between two inferences. Either the bards deliberately substituted Greek for foreign names, or the rulers of the Troad were Greeks. The second alternative, startling as it may appear, seems to us to accord with other evidence and to afford the most satisfactory explanation of the data of the Iliad. If there had been any deep or radical distinctions between the Achæan and Trojan civilisations, it is difficult to see how these could have been completely ignored or successfully concealed by poets who gave such a faithful representation of the topography and evidently were fully acquainted with the character and resources of the enemy. But the strongest argument is the preeminent place which the worship of Athene held at Troy. If that fortress had been under the protection of Apollo or Thracian Ares or Indo-Germanic Zeus, the fact might have little significance; but that a non-Greek people should have adopted such a specifically Greek deity as Pallas Athene for their special patroness is almost incredible. Is there any good reason to resist the simplest and most logical conclusion that Greeks had conquered the Troes and settled in the Troad, and that Mycenæan Troy was a Greek outpost?

According to this view the Trojan War would not lose the significance which Mr Leaf attributes to it, but the historical perspective is somewhat altered. In the first place, Greek enterprise had succeeded in winning possession of the approaches to the Black Sea at a far earlier period than is generally supposed; and this is a new glory in the history of Greek achievement. Nor is it surprising. Mr Leaf has laid stress on the necessity of Euxine commerce to Greece. This was a question which must have been no less insistent in pre-Achæan days; and the Hellespontine regions might well have attracted the ambitions of brave men before Agamemnon. In the second place, the conquest of Troy, apart from its purpose to open up free communication between the Ægean and the Euxine, is seen to be a supplement to the Achæan conquest of Greece. The Achæans have reduced the great Greek states of the peninsula; in attacking Troy they go on to reduce a great Greek state which had established itself in Asia Minor.

Can we conjecture whence the Greek founders of Troy came? Was it possibly from Attica? This would explain the appearance of the Attic Poseidon, Erichthonius, in the genealogy of the Trojan kings, and render it unnecessary to have recourse to the expedient of supposing that the passage where he is mentioned in the Iliad is an interpolation made, in the interest of the Athenians, to give them a sort of title to Sigeum. It would explain too the legend, otherwise curious, that Poseidon aided in building the walls of Troy. It is worth while to notice that Strabo mentions a 'modern' theory that there was some ancient connexion between Troy and Attica.

We may conclude by asking whether the leading motive of the Iliad-defeat suffered by the Achæans in consequence of a quarrel between Agamemnon and Achilles-may have been based on history. Defeat is more inspiring than victory; it makes a deeper impression on the imagination; and the Muse is not satisfied till she has explained it or glorified it. The appearance of Theodoric the Ostrogoth, as Dietrich of Bern, in Burgundian legend, may be taken as an illustration of the powerful influence of defeat. Mr Leaf, who refers to Dietrich more than once, adopts the usual view