which play such a prominent part in the siege, should not have been discovered. The excavations have revealed three gates, situated on the east, south-east, and southwest; and the last of these was blocked up. It is highly improbable that all the gates lay on the southern side of a line drawn from east to south-west. We may assume with virtual certainty that there was at least one other gate, and we may confidently place it on the north-west, where the evidence of the poem would lead us to locate the Scæan Gate. Over the spot where it is reasonable to look for it 'stand great mounds of débris, the spoil of Schliemann's earlier excavations; and until these are removed no certainty can be attained.'

*

It is an interesting question whether any other gate is indicated by Homer. In three passages we read of 'the Dardanian Gates'; and the meaning was discussed in antiquity. Aristarchus contended that this was only another name for the 'Scaæan Gates.' It does not appear why this gate should have been called Dardanian, even supposing that it had an alternative name. Another view is that Dardanian simply means Trojan, and that the phrase means all or any of the gates. This would suit the sense, but is rightly rejected by Mr Leaf, for Dardanian is never equivalent to Trojan in Homer. Dr Dörpfeld holds that the south-eastern gate is meant, as that which led to Dardania. This interpretation, which is obviously in itself satisfactory, suits one of the passages admirably. When Hera says, 'So long as Achilles came forth to war, the Trojans never ventured even outside the Dardanian Gates,' a reference to the issue which lay furthest from the side of the Greek camp would clearly give most point to her rhetoric. But the other passages are not so easily amenable to this interpretation. When we read that, so often as Hector 'set himself to dart under the well-built walls under the Dardanian Gates, if haply from above they might succour him with darts, so oft will Achilles gain upon him and turn him towards the plain,' Mr Leaf thinks that the meaning 'imperatively required' is any of the gates.' But on his own showing Dardanian cannot mean this; therefore, unless we make the unlikely assumption that

* Iliad, v, 189; XXII, 194 and 413.

the word is corrupt, the poet thought otherwise and designated one particular gate. Now Hector's flight began at the Scæan Gate. He ran southward, then eastward; and the first gate he came to, where he could seek shelter, was the south-eastern, as the south-western had been blocked up. Would it not be natural enough for the poet, with a vivid picture of Hector's course in his mind, to select this gate for his description? In the third passage Priam, when he hears of Hector's death, is described as eager to go forth from the Dardanian Gates to seek Achilles at the ships. Why not the Scæan? asks Mr Leaf. Why should he propose to go from the southeastern gate on the other side of the fortress and leading directly away from the sea?' It is true that, issuing by this gate, he would have some three hundred yards farther to drive. But there is another consideration. The Scæan Gate was used when the troops went forth to fight, but it must have been usual during the siege for those who had to go forth for non-combatant purposes to use the south-eastern gate away from the danger side. That is implied in the words of Hera quoted above. It would, therefore, be perfectly natural for the poet to think, and make Priam think, of this gate, just because his errand was not battle. Nor does this suggestion seem to be overruled by the fact that Priam had previously used the Scean Gate when he went forth to conclude an armistice. On the whole we are inclined to believe that Dr Dörpfeld's theory is right, and that these passages taken together are another indication of the general accuracy of Homer's conception of Troy.

The royal Palace has not been discovered. It was probably situated in the centre of the castle on ground which has been entirely levelled. It is possible that the site of the temple of Athene has been found. While the houses are built on the same plan, with megaron and prodomos, there is one which is distinguished by the singular feature of a line of three columns down the centre, an arrangement which existed in a Greek temple at neighbouring Neandria. If this was a temple, as has been conjectured, it may have been that on which Hecuba and the Trojan women laid their peploi on the knees of Athene in the sixth Book of the Iliad. Mr Leaf does not commit himself very clearly, but he seems to imply his

belief that the worship of Athene was Trojan. For of course it might be maintained that this is a Homeric anachronism, and that Athene was first worshipped at Troy by the inhabitants of the succeeding settlement, who may have been Greek.'

This is an important point and we ought to be clear about it. The curious story of the Locrian tribute, to which Mr Leaf has devoted some instructive pages, seems sufficient to establish the cult of Athene in the city of Priam. The crime of the Locrian Aias, who violated Cassandra, the priestess of Athene, tearing her from the sanctuary, is familiar. Plague and famine broke out in Locris; and the people were told that the visitation was due to the wrath of the goddess, and that they must send two maidens every year for a thousand years to serve in her temple at Troy in order to propitiate her. 'So far,' says Mr Leaf, 'is legend only; but the hundred noble families of Locris actually sent two maidens chosen by lot every year, and this practice was still maintained at the end of the fourth century B.C., as is proved by an inscription. Mr Leaf shows that the practice had begun not later than 800 B.C. He assumes that Greek settlers on this site (in the later period of the Seventh City) felt it necessary to make atonement to the deities whose shrines had been made desolate by Greeks; and 'nothing is more likely than that Delphi should make reparation for the legendary sin an obligation on any colony in Troy itself.' This seems to assume that Athene was worshipped in Mycenæan Troy; but we are inclined to be more conservative than Mr Leaf, who regards the tale of Aias as unhistorical. Why should such a legend be invented? If the incident is true, it explains the Locrian tribute and it also explains itself. If it was an invention, its only imaginable motive was to explain the Locrian tribute. But in that case the true origin of the Locrian tribute remains unexplained. If we believe, like Mr Leaf, in the Trojan War and the Achæan capture of Troy, there is no reason for refusing to believe that the incident actually occurred. Unfortunately for Cassandra, it was one of the things which were likely to happen in the sacking of the city. If this argument is reasonable, it supports the conclusion, which otherwise seems probable, that Athene had a temple in the Troy of Priam.

When he comes to examine the Homeric geography of the Troad and Asia Minor, Mr Leaf is led by logical steps to a theory which forms his most original and striking contribution to the elucidation of the Trojan problem. His main text is the Catalogue of the Trojan forces in the Second Book. He is able to show that the geography of the Catalogue is in accordance with the rest of the Iliad, and that it implies conditions completely different from those which prevailed when the Iliad assumed its present form. These conditions suggest an hypothesis which enables him to account both for the rise of the Trojan power and for the outbreak of the Trojan War.

The overlordship of Priam seems to have extended over the Troad itself, westward to the river Esepus, beyond which was the country of the Mysians, and southwards to the Gulf of Adramyttium. Here Mr Leaf has done much to clear up the geography of the southern Troad, which he shows convincingly to have been a Pelasgian confederacy, to which the worship of Apollo Smintheus was common. He reconstructs from various passages in the Iliad the great foray of Achilles which was directed against the cities of this confederacy. This raid, in which Chryseis and Briseis were captured, was the immediate prologue to the action of the Iliad. The allies beyond the Troad who are enumerated in the Catalogue fall into four groups- the Pæonians and Thracians in the north; the Paphlagonians in the far east; the Phrygians and Mysians in the near east; the Mæonians, Carians, and Lycians in the south. Here, says Mr Leaf, we have four radial lines which represent four trade routes leading straight to what were the chief centres of trade in the early days of Greek colonisation, to Miletus, Amphipolis, Cyzicus, Sinope.' The inference is that Troy was a great commercial centre, and the question arises why? For no spot would seem less marked out by nature for commercial prosperity than the plain of Hissarlik. With its marshes and malaria, it was a poor place compared with other plains in the Troad. Moreover

'there is no natural harbour in the district. Troy cannot therefore have thriven upon her over-sea commerce, or its

close relation, piracy. Troy has indeed two roadsteads, one to the north of the Hellespont, the other, Besika Bay, to the west; but both of them are exposed anchorages, offering no safe shelter in gales.'

How then did a strong and wealthy power arise in such an unattractive situation, suitable neither for production nor for commerce nor for plunder? The secret, according to Mr Leaf, is to be found in the conformation of the Hellespont and the prevalence of certain winds in the eastern Mediterranean.

'It is easy to see that the condition which is needed in order that Troy may be an important centre of commerce is that the Hellespont should be closed to the ships of the Ægean Sea. When this is the case, the Trojan plain becomes of necessity the natural meeting place for the trade of the Ægean and the Euxine. . . . The passage of the Hellespont is easily closed against sailing ships by those who hold the land. The dominant factor in the navigation of all the eastern Mediterranean is the prevalence throughout the summer of the Etesian winds, blowing from N.W., N., or N.E., often with great violence for many days together. Any sailor making for the Propontis must perforce reckon on a delay at the mouth of the Hellespont, almost certainly for some days, perhaps for a fortnight or so. In early times, and indeed so long as galvanised iron tanks remained unknown, the water supply was a vital question for all navigation. Only a poor supply could be carried in the heavy earthenware jars on which the Greeks depended; and so it was that a delay of even two or three days wind-bound on a coast where the water supply was in hostile hands, was a matter of life and death. . . . The natural supply of water for ships making the passage of the Straits is of course from the Scamander itself. This is easily defended; there is no other permanent stream for several miles. . . . There is also good water to be had at Besika Bay, a fact which has on more than one occasion proved serviceable to the British fleet. But Troy is so placed that it can easily command this also. A garrison in the castle could easily keep watch over both sources by stationing at them detachments sufficient to oppose any unauthorised landing by the crew of a merchant ship' ('Troy,' pp. 261–2).

Thus the lord of the Trojan plain had it in his power to exclude from the Euxine the merchants of the Ægean,