While the events just described were taking place on land, a somewhat desultory naval warfare was being waged in the channel north of Eubœa, which, though in itself indecisive, was not without influence on the ultimate issue of the war. The defence of the channel had been rightly regarded as an absolutely necessary adjunct to that of the pass; and the resolution to send troops to Thermopylæ had been accompanied by the resolution to send to the North Euripus all the ships available. Even Herodotus recognizes the intimacy of the connection between the two positions. When visiting Thermopylæ he can hardly have failed to notice the relation between them.

Speaking of it he says: “These places are near one another, so that the forces stationed at them could communicate with one another.”

Even from a point at Thermopylæ only a hundred and fifty feet above the Malian plain it is possible to see right down this northern channel, and to distinguish in the distance part of the outline of the island of Skiathos.

The historian treats the Artemisium story as an episode in itself, though he supplies means by which it may be chronologically connected with the Thermopylæ narrative. This treatment is justified by the fact that, though there was so intimate a connection between the two positions from a strategical point of view, the actual fighting which occurred at them was of a quite independent character in so far as he could judge of it. But there is also another reason which probably had still greater weight in determining the method which he employed. SPARTAN AND ATHENIAN SOURCES. The story of Thermopylæ is plainly of Spartan origin from beginning to end; that of Artemisium is drawn from Athenian sources. There is a further difference between them. The story of Thermopylæ is, in respect to motive, a popularized tale of official origin, for which the authorities at Sparta were responsible. That of Artemisium is derived mainly, like some other parts of Herodotus’ military history, from some one who was present at the engagements, but was not in a position to know the designs of those in command. The tale is too confused for it to be possible to suppose that it is founded largely on official documents or on official information of any kind. The reasons given for important movements are such as would be current in the gossip of the fleet, coloured by the evident intention to bring into high relief the greatness of the service which Athens did to the Greek cause at this time by saving the situation on the Euripus. The impression which is so intentionally given was not indeed a false one. The Athenians did save the situation at Artemisium in so far as it could be saved. By so doing they contributed not a little to the solidarity of the Greek resistance.

The method which Herodotus employed to bring the two series of events into connection with one another is natural; but the result is curious. His narrative of this part of the war is in the form of a twofold diary of the incidents at Thermopylæ and at Artemisium respectively. There are two points of contact between the two diaries—at their beginning, the departure from Therma, and at their end, the disaster at Thermopylæ. It is notable, however, that there is a discrepancy of two days between them. It is probable that two days have somehow been omitted from the Artemisium diary. A close examination of its narrative makes it evident that the historian, either owing to his own miscalculation, or to some error in his source of information, has attributed to one period of twenty-four hours events which could not conceivably have happened within so brief a space of time. That it is a case of pure miscalculation, either on his own part or on that of his authority, H. viii. 15. is further evident from his own statement that the naval battles at Artemisium took place on the same days as the battles in the pass at Thermopylæ; in other words, that the events of the eighteenth, nineteenth, and twentieth days of the Thermopylæ narrative were synchronous with those of the sixteenth, seventeenth, and eighteenth days in the record of the Artemisium journal as it stands in his work.129

The fleet which the Greeks despatched to Artemisium was a large one, composed of the following contingents:⁠—

Athenians, 127 triremes, part of the crews of which were Platæans; Corinthians, 40 triremes; Megareans, 20 triremes; Chalkidians, (probably Athenian kleruchs from Chalkis,) manned 20 triremes provided by Athens; Æginetans, 18 triremes; Sikyonians, 12 triremes; Lacedæmonians, 10 triremes; Epidaurians, 8 triremes; Eretrians, 7 triremes; Trœzenians, 5 triremes; Styreans, 2 triremes; Keians, 2 triremes and 2 pentekonters: Opuntian Locrians, 7 pentekonters.

The total number Herodotus gives correctly as 271. The commander of the fleet was the Spartan Eurybiades. ATHENS AND ATHENIAN SOURCES. The story goes that, before the embassy was sent to Sicily, it had been arranged that Athens should have the supreme command at sea, but that the Peloponnesian allies opposed this arrangement even to the extent that they threatened to withdraw their contingent altogether if it were allowed to stand. And so, says Herodotus, the Athenians, unwilling to sacrifice the national cause, with a laudable patriotism waived their claim. The refusal of the Peloponnesians was not unnatural. It was one thing to submit to the hegemony of the ancient and respectable Sparta; it was another thing to place themselves under the leadership of this pushing parvenu state, whose rise to importance had been so extremely rapid. On the other hand, the magnanimous attitude of Athens was based, no doubt, on the simple arithmetical calculation that 270 ships were better than 150. Herodotus implies that there had been a change in the attitude of the allies. The profound dissatisfaction which they felt with the Northern policy may easily account for this. At the time of the embassy to Sicily the question of the locality of the defence had probably not been raised in a highly controversial form.

In this part of the Artemisium narrative the Athenian bias is as strongly marked as elsewhere: it contrasts, however, with the major portion of the story in the fact that it seems to rest upon an official basis. The details of the numbers of the various contingents, and the correct sum-total of the whole,—a trait lacking at times in Herodotus’ history,—render it unlikely that the information can have been derived from mere tradition, or from personal recollection.

The fleet of Xerxes remained at Therma for twelve days after the army had departed. It had now the most difficult and dangerous part of its voyage to perform. From Therma to the Pagasætic Gulf there is no harbour into which it could put; and it is probable that the intention was to perform the passage in one continuous stretch.

The fast-sailing vessels were sent forward as scouts towards the island Skiathos, which lies off the mouth of the Euripus. H. vii. 179. These vessels came across three Greek ships, Trœzenian, Æginetan, and Attic vessels respectively, engaged in a similar duty off the mouth of the Peneius river. H. vii. 182. When the Persian vessels came into sight, the Greek ships fled. Not merely were they outnumbered, but also up to this time the Greeks had had no experience which could give them confidence in their own naval equipment when matched against that of the Persian. H. vii. 180, 181. The Trœzenian and Æginetan vessels were taken with their crews; the Athenian ship was run aground at the mouth of the Peneius. Those aboard her escaped to shore, and finally, after what must have been an adventurous journey, made their way to Athens. H. vii. 179. 183. Apparently on the same day, the main body of the fleet left Therma, and, after sailing the whole day through, “arrived at Sepias and the strand between the city of Kasthanea and the Sepiad Cape, in Magnesia.”

On this day the Athenian fleet was at Artemisium. The promontory of that name, crowned by the temple of Artemis, was the north-east point of the island of Eubœa. The actual station of the fleet must have been at least ten miles from the promontory, well within the strait, at a point west of the mouth of the Pagasætic Gulf, where the fair way is much narrower than in the outer part of the channel. H. vii. 176. The name “Artemisian shore” seems to have been applicable to this part of the coast of Eubœa, as well as to the immediate neighbourhood of the promontory.

The capture of the Greek outlook vessels was reported to the Greek fleet at Artemisium by means of fire signals from Skiathos.130

The possibilities of the case preclude the supposition that the Greek fleet received the news sufficiently early for them to move from Artemisium on that day. It therefore becomes a serious historical question as to what interpretation is to be put upon the account which Herodotus gives of the next step taken by the fleet. He says, “On hearing this news the Greeks were frightened, and removed from their anchorage at Artemisium to Chalkis, with intent to guard the Euripus, leaving watchers posted on the heights of Eubœa.”

There are almost overwhelming reasons for doubting the truth of the motives of this story; though, as is so often the case in Herodotus’ narrative, the facts appear to be true. It is probable that the Greek fleet did move towards, if not to, Chalkis, and that watchers were posted on the heights of Eubœa. But the reason given for this change of station is that the Greeks were panic-stricken at the capture of their three scouting vessels. This reason is inadequate and incredible.

1. The Greeks had made up their minds to face the Persians at Artemisium. Is it credible that so small a disaster should have led to the adoption of a resolution, the effect of which, under the circumstances, must have deranged the whole plan of campaign?

2. If such a panic had taken place for a reason so insignificant, is it probable that any human power could have induced the Greeks to return a few days later to their former station?

3. Is it credible that they should have retreated without warning the garrison at Thermopylæ which would in that case have been sacrificed?

4. If that garrison had been warned, would it not have retreated from that advanced post in the same way that the army had withdrawn from Tempe?

5. It is extremely unlikely that the withdrawal was made at night. It is almost certain that it must have been made on the morning after the news had been received, by which time the storm which burst upon the Persian fleet at the Sepiad strand had already begun.

6. The motive alleged is in the highest degree suspicious, because all through the narrative the services of the Athenians in keeping the fleet at Artemisium are emphasized in the most marked manner.

It is probable that the motive given is a pure fabrication invented at Athens and current at Athens in after-time. It is also probable that the real reason for the withdrawal of the Greek fleet was the impossibility of remaining in the North Euripus during the storm. Such storms spring up with astonishing suddenness in these seas, and give little or no warning of their approach; and it is probable that the Greek fleet, caught unprepared, had to run for it to the inner strait, where, once round the corner of Eubœa, it would be sheltered by the mountains which form the great precipices on the Euripus north of Chalkis. Whether it went to Chalkis or not, it is, of course, impossible to say. It is probable it did not do so, for Mount Kandili131 ceases before Chalkis is reached, and the narrow part of the Strait is exposed to the east and north-east. There is no difficulty in identifying the nature of the storm. It blew on shore on the Sepiad strand, and therefore must have been one of those E.N.E. gales from the Black Sea, which are the most violent and most dangerous of the storm winds of the Ægean. They may occur at any time of year. Such a storm would blow right down the north bend of the Euripus.

The army at Thermopylæ must have known of the retirement of the fleet; but it must have known also that there was every intention of returning to the former station so soon as the weather allowed. For that reason it held its ground.

The adventures of the ten Persian ships did not end with their capture of the Athenian vessels. Three of them came to grief on the rock called Murmex, which is in the fairway just at the entrance of the channel.

On the evening of the twelfth day of the journal the position appears to be as follows: the Greek fleet was at Artemisium; the Persian at the Sepiad strand outside the strait; the Persian scouting vessels, or, rather, the seven survivors of them, H. vii. 183. were probably engaged in erecting a pillar on the Murmex rock as a danger mark to the rest of the fleet.

The Sepiad shore was not sufficiently extensive for it to be possible to draw up the whole of the Persian fleet upon it. A large number of the vessels were obliged to remain at anchor, with their bows facing seawards, in eight lines. 13th day. These were surprised at dawn by the storm. It arose suddenly, the sea having previously been quite calm. The wind was from the east or E.N.E., known in that region as the “Hellespontian wind,” not less known, nor less obnoxious to sailors at the present day than it was in the days of Herodotus. The destruction wrought in the fleet was enormous. Four hundred warships and an unknown number of transports are said to have perished. The numbers are no doubt greatly exaggerated; but the loss relatively to the size of the fleet seems to have been very considerable. Such were the events of the thirteenth day at the Sepiad strand. The Greek fleet, driven from the North Euripus on that morning by the storm, took refuge in the inner channel,—“at Chalkis,”—so says Herodotus.132

The storm lasted during the fourteenth and fifteenth days, the fleets remaining at their respective stations.

The events of the sixteenth day are of special importance. They are crowded together in the narrative in a way that raises the suspicion that the chronological error in the story is to be sought in this part of it.

The storm had come to an end on the fourth day; but on the day following its commencement the Greeks received news of the disaster at the Sepiad strand from the watchers posted on the heights of the Eubœa.

“When the Greeks heard this, after offering prayer to their preserver Poseidon, and pouring libation, they hastened straightway to Artemisium, hoping that but few ships would oppose them.” There is manifestly some mistake here. The Greek fleet could not possibly have made its way back to Artemisium in the face of the storm. 16th day. It must have waited until the storm had blown itself out; that is to say, it could not have returned to its former station until the sixteenth day. The tale as it stands supplies a motive for the return of the fleet such as will fit in with the main motive of the story.

“They came for a second time to Artemisium and took up their station there, and from that day to this have been accustomed to address Poseidon by the name of ‘saviour’ then given to him. The barbarian, when the wind dropped and the sea went down, dragged the vessels down into the water and sailed along the coast round the cape of Magnesia, into the gulf leading to Pagasæ. There is a place in this Gulf of Magnesia where it is said that Herakles was left by Jason and his comrades when he had been sent from the Argo to fetch water, when they were sailing to Aias in Colchis after the golden fleece; for their intention was, after watering there, to put out to sea. From this circumstance the place got its name of Aphetæ. Here the fleet of Xerxes came to anchor.”

It is fairly clear from the description that Aphetæ must have been situated at the extremity of the long curved peninsula which extends west from the south end of Magnesia, enclosing the Pagasætic Gulf on the south.

The only other event which can be reasonably ascribed to this sixteenth day is the capture by the Greek fleet of fifteen Persian vessels which were belated in starting for Aphetæ. It is not difficult to demonstrate, on Herodotus’ own showing, that their capture cannot have taken place until well on in the afternoon.

The tale is as follows:

The storm had not ceased until that very morning. It is quite certain that the battered Persian fleet must have required some hours of preparation before it could put out from the Sepiad strand; and it did not reach Aphetæ until the early afternoon. The fifteen vessels must have come up some hours later, well on in the afternoon, since the main fleet had evidently disappeared into the Pagasætic Gulf before they entered the channel; otherwise they could not have made the mistake they did make. The story of their capture shows that on the afternoon of that day the Greek fleet was already back at Artemisium. Had it come up from Chalkis that morning? It is not impossible that it should have done so; but it is improbable.133

“So the Persian fleet, with the exception of the fifteen vessels which, as I have said, Sandokes commanded, came to Aphetæ.”

The composite character of the sources of the historian’s narrative is well illustrated by the next episode in the strange story. It is of the nature of a by-plot in the drama inserted with a view to illustrate not merely the services of the Athenians in saving the situation, but also the twofold character of the man who now becomes for the first time prominent in the story. H. viii 4. In brief the tale is that the fleet, after returning to Artemisium, discovered that the disaster at the Sepiad strand had by no means been so great as had been imagined. So alarmed were the Greeks at this, that “they discussed whether they should retreat from Artemisium to Middle Greece.” The Eubœans, who thus saw themselves likely to be suddenly deserted without the possibility of providing for the safety of their children and households, implored Eurybiades to allow them time to remove them to a place of security. Finding they could make no impression upon him, they resorted to Themistocles, to whom they offered a bribe of thirty talents, if he would take measures to keep the fleet where it was. Themistocles undertook to do so, and proceeded to bribe Eurybiades with a gift of five talents, and Adeimantos, the Corinthian admiral, with three talents.

The tale excites the strongest suspicion of having been an invention of after-time, concocted by political opponents of Themistocles. Herodotus’ whole attitude to Themistocles is suspicious. He is quite unable to suppress the greatness of the services which he rendered to the Greek cause; but he is ever ready to discount the value of those services either by tales discreditable to the personal character of the man who rendered them, H. vii. 144. or by attributing to him motives other than those which really prompted his action.

Nor do the details of the story, when taken into consideration, tend to make it appear more probable. No doubt the Peloponnesian members of the fleet seized every possible occasion for urging such a withdrawal. It must have been known to them that the force sent to Thermopylæ did not represent any real effort on their part. It could hardly be expected that they should display any large amount of zeal in carrying out a policy of which they wholly disapproved, and in the execution of which their own authorities were so manifestly half-hearted. They knew that, if they left Artemisium, the army must leave Thermopylæ; and they were no doubt convinced that that was the very best thing which could possibly happen. But on the practical question of withdrawal, other considerations must have influenced those in command. Unless the retreat from Thermopylæ were begun many hours before the withdrawal of the fleet, the army must be caught in a trap. With a light-armed enemy in vastly superior numbers in hot pursuit, the retreat could not have been rapid, and the Persian fleet would have had ample time to land troops in such a position as would make it possible to cut the line of it.

There is no trace whatever in the Thermopylæ story of a message having been received from the fleet advising the garrison of its intended withdrawal. That may be attributed to the incompleteness of the story itself. This explanation, however, can hardly be considered satisfactory in view of the fact that, had anything of the kind occurred, such an incident would hardly have been omitted from a version the main object of which was to depict the disaster as an act of deliberate self-sacrifice on the part of Leonidas.

The decisive factor in the situation was that, bribed or not, Eurybiades and Adeimantos could not possibly leave the Thermopylæ garrison in the lurch.134 The real significance in this, as in the other similar stories in Herodotus relating to this time, is that it brings into relief the existence of a strong party within the fleet which was wholly opposed to the Northern policy. It also for the first time indicates the part which Corinth seems’ to have played in the war as the leader of this party of opposition to the Athenian policy. Thus far it is reliable. The rest of the tale is, as has been said, suspicious to the last degree,—suspicious, in what may be called its local colouring,—suspicious, owing to the appalling recklessness of the assertions which Greek politicians did not hesitate to make respecting their political opponents.

Diodorus’ version of Artemisium presents so many marked differences to that of Herodotus, that it is impossible to suppose that it is either drawn from Herodotus, or even derived from a common source. There is absolutely no hint of bribery in it, nor, indeed, is there any reference to the restiveness of the Peloponnesian contingent. What is emphasized is the fact that Themistocles’ personal influence in the fleet, and especially with Eurybiades, was very great, and that he was the de facto commander. There is no tendency to belittle the services of Athens, though her claims to merit are put upon a different basis.

It is probable, then, that the Greek fleet, after its return to Artemisium on that sixteenth day, did not face the situation with a unanimous determination to make a stubborn fight of it. It would have been surprising had unanimity prevailed, since the circumstances as they stood, and as they were intended to stand, at Thermopylæ cannot possibly have been a secret to members of the Peloponnesian contingents. Still, so long as Leonidas remained there, the strait had to be defended.

The Persians at Aphetæ were animated by very different feelings. H. viii. 6. The Greek fleet was not to be defeated merely, but to be captured. 17th day. It was determined to despatch a large squadron, H. viii. 7. two hundred in number, to circumnavigate Eubœa, and to block the South Euripus.135

Exaggerated as are the numbers given in Herodotus, the mere fact that the Persian commanders ventured thus to divide their fleet in face of an enemy with 271 ships, together with the comparative ill-success with which the Greeks fought with that part of the fleet which remained at Aphetæ, shows that the Persians must, even after the disaster on the Sepiad strand, have had a very large number of vessels at their disposal.

Herodotus does not say, but he certainly implies, H. viii. 7. ad fin. that this flying squadron was despatched on the very day of the arrival at Aphetæ. It is not necessary to insist that this was, under the circumstances, hardly possible. The fleet had been terribly knocked about by the storm. The storm itself had only ceased that very morning. There had been no time to refit before leaving for Aphetæ; and the fleet had only reached that anchorage early in the afternoon of that sixteenth day. H. viii. 7, ad init. These two hundred vessels, moreover, were evidently despatched by daylight.136

It is here, no doubt, that one of those missing days has been omitted from Herodotus’ narrative of Artemisium. The despatch of the squadron round Eubœa must have taken place on the morning of the seventeenth day.

The one fear of the Persians at this time was that the Greek fleet would run away. They, consequently, adopted the stratagem of sending the squadron of two hundred vessels by the channel north of Skiathos, so as to give the impression that it was sailing to Therma or elsewhere. It was an effective manœuvre, because the island completely fills in the sea horizon of the mouth of the channel.

On the next day137 they reviewed their fleet at Aphetæ, but otherwise lay quiet, intending to wait for news of the encircling squadron.

Of the destination of the two hundred the Greeks do not appear to have had any suspicion whatever, until Skyllias the diver, deserting the Persian fleet, made his way across the strait and gave them information. The utter confusion of this part of the Artemisium story renders it very difficult to arrive at anything resembling a certain conclusion as to what actually took place in consequence of the receipt of this news. It is hardly conceivable that those in command of the Greek fleet should have been quite unprepared for the eventuality of a squadron being sent round to intercept their retreat by Chalkis. There is no direct evidence on this point; the slight evidence which does exist is of a circumstantial character. H. viii. 14. After describing the disaster which overtook the two hundred vessels, Herodotus says that the news of it was brought to the fleet at Artemisium by a squadron of fifty-three Attic vessels, or, at any rate,—for his language is not quite explicit,—that the news arrived with that squadron. This is the first mention of this body of ships. It seems highly probable that it had been at Chalkis, guarding the narrow passage there, and that, on receiving news of the disaster to the Persian flying squadron, it went to join the rest of the fleet at Artemisium.

The escape of Skyllias is asserted to have been made during the review, and the consultation among the Greeks as to what course they must adopt in consequence of his information was held on the day of the first engagement.138

Herodotus reports that they determined to remain where they were for that day and the greater part of the following night, and then, after midnight, to go and meet the ships coming up through the south channel. It is quite intelligible that, if the Greeks had not provided for the defence of the narrows at Chalkis, their position was such as to cause alarm, and to demand desperate measures.

But the probability is that there never was any intention on the part of the Greek commanders to leave the station at Artemisium until some blow had been inflicted on the fleet at Aphetæ.139

The Attic vessels were probably guarding the strait, which at Chalkis was so narrow as to be easily defensible. Under no circumstances could the fleet at Aphetæ have been left free to land troops behind Thermopylæ.140

The subsequent action of the Greek commanders is in accord with such a situation and such a design. They had not hitherto ventured to attack the Persian. They were now forced to do so. Not until the fleet at Aphetæ had been rendered incapable of making a movement down the Euripus could the Greek fleet venture to meet the danger from the south. And so the first attack took place that very night. Herodotus’ account of the action is a curious one,—that of a man who had heard talk of certain naval technicalities without understanding them. These very Greeks, who have been described as having, a few days before, fled to Chalkis in consequence of the capture of three of their vessels, and as having contemplated flight to mid Greece at the mere sight of the Persian fleet, are now represented as making a quasi-experimental attack on the foe they so much dreaded, from motives of purely professional interest. “They wished to make trial of the Persian mode of fighting, and to practise the manœuvre of cutting the line.” 141

From the point of view of military history, the account of the first engagement, as given by Diodorus, conveys a much stronger impression of accuracy than that of Herodotus. H. xi. 12. Both represent the Greeks as the attacking party; but Diodorus gives an excellent tactical motive for their taking the offensive. A council of war being held, all the commanders save Themistocles were in favour of remaining on the defensive. He, however, persuaded them to adopt his plan of the offensive by pointing out to them the advantage they would enjoy as the attacking party, in that they, with their whole force prepared for battle, would be able to choose their own time for attack, and must take unawares an enemy whose fleet was not concentrated at one anchorage.

This account is strongly supported by the fact that there is no single harbour at that extremity of the Magnesian peninsula which could accommodate more than a fraction of the Persian fleet, so that it would necessarily be distributed over a considerable extent of coast-line.

On the general course of the engagement the two accounts are in agreement. The Greeks, as might be expected, gained at first some considerable success;—Herodotus says they took thirty ships; but, when the whole of the enemy’s fleet had collected, the combat became obstinate and indecisive, until the fall of night put an end to it.

It was on the evening of this day that the second of those storms took place which were destined to be so fatal to the Persian fleet. STORM AT THE HOLLOWS OF EUBŒA. The quarter from which it came is not expressly stated by Herodotus; but his accounts of its effects on the north and south of Eubœa respectively leave no doubt that it was a gale from the south or S.S.W.142 At Aphetæ it caused considerable alarm but little damage, the only effect being that the wreckage from the battle of the afternoon was thrust in upon the Persian fleet.

In the South Euripus there was a very different tale to tell. The Persian flying squadron, which had started from Aphetæ the day before, appears to have been caught by the gale immediately after it had rounded the southern Cape of Eubœa.143 The result was a disaster hardly less great than that which had happened at the Sepiad strand. The whole two hundred were driven into those bays at the south end of the west coast of Eubœa, which the ancients knew under the name of the “Hollows.” Not a ship escaped. The destruction was complete.144

The effect of this disaster on the general course of the war can hardly be over-estimated. It was not merely that the numerical loss was great, but the position of the fleet at Artemisium must have become exceedingly critical, had this squadron occupied the channel south of Chalkis. Under such circumstances it is probable that the decisive battle of the naval campaign would have been fought near that place.145

It was probably early in the night of the eighteenth day that the disaster at the Hollows took place. The report of it was promptly carried to Chalkis, at which place the fifty-three Attic vessels must have ridden out the storm, since a gale from the south raises a very ugly sea in the broader parts of the Euripus. It is evident that the Greeks had established a regular signalling system on this coast; and it is probable that it was by this means that the news reached the squadron at Chalkis, which immediately started for Artemisium. H. viii. 14. Herodotus’ language leaves it a somewhat open question as to whether these ships brought the news to the fleet. All that is actually stated is that the news arrived at Artemisium about the same time as the fifty-three.

The Persian fleet, ignorant, no doubt, of the disaster in South Eubœa, intended to bide quiet until some report of the two hundred reached it. For the earlier part of the day it remained undisturbed. The Greeks had probably found out to their dismay on the previous evening that they were not strong enough to tackle the naval force which remained at Aphetæ with any hope of decisive success. H. viii. 14. But the arrival of the fifty-three Attic vessels, which can hardly have taken place before midday, put new heart into them; and that afternoon they attacked the Cilician contingent of the Persian fleet, inflicting considerable damage.146

The Persian position, both at Thermopylæ and Artemisium, was at this moment very precarious. The army had conspicuously failed to force the pass, and the fleet had, so far, confined itself to the defensive. Cf. H. viii. 15, ad init. Doubtless, urgent messages were coming from Thermopylæ to Aphetæ, commanding that the strait should be forced at any price. 20th day. The Persian commanders, therefore, afraid to defer longer the attack, took the offensive. It may well be that the disadvantages of the position at Aphetæ from a defensive point of view, which had been clearly demonstrated in the engagements of the two previous days, had something to do with determining their course of action. The battle which ensued was an obstinate one. It certainly was not favourable to the Greeks, even less favourable, perhaps, than is represented. H. viii. 16. The losses on both sides were large, those of the Persians being the greater. It seems, however, that, over and above the vessels actually destroyed, H. viii. 18. a large number were considerably damaged, and that the Athenian contingent suffered especially in this respect. H. viii. 19, 20. The account which Herodotus gives of the events which took place after the battle leaves it uncertain whether he considered that the subsequent retirement from Artemisium was due primarily to the losses incurred in the previous engagement, or to the receipt of the news of the disaster at Thermopylæ. Diod. xi. 13. Diodorus mentions the losses, but ascribes the retreat entirely to the news from Thermopylæ. That disaster must have taken place early in the afternoon: the news of it must have reached the fleet at Artemisium, by means of the despatch-boat, very early in the evening. H. viii. 21. According to Herodotus, the retreat had been discussed before the despatch-boat arrived.147 That it had been discussed may well have been the case; but that it had been decided, and, above all, that Themistocles had consented, to retire that very night, before the receipt of the news from Thermopylæ, is doubtless untrue. Herodotus’ assertion incurs all the more suspicion from the fact that he uses it as a peg whereon to hang a tale illustrative of the disaster which is sure to befall those who refuse to obey the advice of oracles. The Eubœans, he says, had disregarded an oracle of Bakis, which had urged them to place their sheep in security, in view of the coming war, and were now, on the approaching retirement of the Greek fleet, in a position of extreme difficulty as to the best means of saving their property. Themistocles’ proposal for dealing with the situation was simple and effective. He said, “Kill the sheep: kindle fires: roast them and put them on board, since it is better that they fall into our hands than into those of the enemy. As to getting the fleet away in safety, I will look to that.” He also hinted that he had a plan, which he did not then disclose, for detaching the Ionians and Carians from the Persian fleet. So the other Greek commanders looked to the sheep and the fires, while he devised measures for retreat.

No tale could be more peculiarly illustrative of the wide distinction which must be drawn between the facts of history as stated by Herodotus and the motives which he propounds for certain courses of action.

It is doubtless perfectly true that a council of war was held that evening, at which it was decided to retire from Artemisium; and that Themistocles concurred in that decision. But, on the question of motive, it may be taken as almost equally certain that this decision was not come to in consequence of what had happened in the engagement of that afternoon, but by reason of the news which had come from Thermopylæ, which rendered the further presence of the fleet at Artemisium not merely unnecessary, but, from the point of view of the Greek commanders, unadvisable. In actual fact it may be regarded as exceedingly doubtful whether Xerxes could have advanced southwards without the aid of the fleet commissariat; but the idea uppermost in the minds of the Greek commanders would be the extreme danger of such a move on his part.

It is not difficult to see that the building of the fires that evening on the Artemisium shore had probably little if any connection with the provisioning of the fleet, but was part of the plan by which Themistocles proposed to secure a safe withdrawal from the position. The Persians were evidently to be deluded by the idea that the Greeks had no intention of deserting their station; and the ruse employed to attain this end would naturally suggest itself. It has been again and again employed under similar circumstances.

The retreat began early in the night, the ships taking their departure in order according to their position at the anchorage. H. viii. 21. The fact that the Corinthians led the way, and that the Athenians were last to start, may merely signify that they formed respectively the extreme left and right of the line; but, in view of the attitude of the former at this time, it may be suspected that Herodotus had special motives for mentioning the order of retirement.

Themistocles, with a fast sailing division, formed the rearguard. He did not, apparently, intend so much to cover the retreat, as to visit the various watering-places on the shores of the Euripus, and to affix there a quasi-manifesto to the Ionian Greeks in the Persian fleet, calling upon them to abstain, in so far as possible, from attacking their fellow-countrymen. H. viii. 22. His object was, says Herodotus, either to detach the Ionians from their allegiance to Xerxes, or to make the king suspicious of their loyalty.

The Persians did not attempt to pursue the Greek fleet. «H. viii. 23.» The latter had, indeed, got a long start before its departure was discovered. A Greek from North Eubœa brought the news to Aphetæ; but his information was distrusted, and ships were sent across the strait to discover whether it was true or not. This must have taken place during the night, for it was at sunrise that the Persian fleet, after receiving the report of the scouting vessels, moved across the strait to Artemisium. At midday the Persians moved to Histiæa, where they appear to have spent some time in ravaging the district in its neighbourhood.

It is probable that the news of the disaster in the South Euripus deterred the Persians from any attempt at pursuit. Nothing is said as to how or when the news reached them; but it may be regarded as certain that the tidings, which had arrived at Artemisium twenty-four hours before, had in the interval made their way across the strait.

It was now possible for the Persian fleet to communicate by sea with the captors of Thermopylæ; and it may easily be imagined that no time was lost in sending supplies to the army. In connection with these communications Herodotus relates a tale to the effect that Xerxes invited the sailors to view the scene of battle, and took certain ridiculously ineffective measures to conceal from them the magnitude of the losses suffered. That the king did attempt to conceal the greatness of the price he had paid for the capture of the pass may well have been the case; but whether Greek tradition, with its marked tendency to bring into high relief the childishness and foolishness of the Oriental monarch, truly represented the measures taken to attain this end, may well be doubted.

In Herodotus’ story of the great war the tale of Artemisium is perhaps the least satisfactory of all the detailed accounts of the various acts of the drama. It is not merely complicated by a chronological error of considerable magnitude, the effect of which is to render the most important part of the story, as it stands, incredible; but it is very seriously distorted, from a historical point of view, by the addition of material of a more than doubtful character, inserted with intent to heighten the effect of the services of Athens at this critical time. But the chronological error does not destroy the historical value of his narrative, though it impairs it. Had Herodotus told his tale after his usual manner, the error might not have crept into it. But just in this part of his history he has tried to treat events in a somewhat more business-like fashion than usual, and the attempt has not met with complete success. Nature had not made him an arithmetician; and even had it done so, his business-like method is sufficiently unbusinesslike to render his mistake obscure until the story has been analyzed.

How far he is personally responsible for the Athenian bias of his version, it is not possible to say at the present day. HERODOTUS’ ACCOUNT OF ARTEMISIUM. He was, no doubt, sufficiently philo-Athenian to wish to place the undoubtedly great services of Athens to the national cause at this time in the highest possible relief; and he was sufficiently ignorant of strategical considerations to render it probable that he could not detect in the version current at Athens those fictitious additions, whose falsity must have become apparent to any one who could appreciate the main lines of the strategy of this part of the war. Had he understood not merely that there was a connection between Thermopylæ and Artemisium, but that connection was of so intimate a nature as to render the maintenance of the pass absolutely dependent on the maintenance of the strait, he would have treated with suspicion those repeated assertions of an intention on the part of the commanders of the fleet to withdraw from their station, or would have reduced them to a position of far less prominence in the story, as being, what they no doubt were, mere indications of the spirit which animated the irresponsible mass of those who formed the crews of the Peloponnesian contingents.

And yet, in so far as can now be judged, his tale of Artemisium,—freed from a chronological error for the correction of which the author supplies the means, and with the essentially Athenian portions of it reduced to their proper relief in the story,—is a reliable history of this important period of the Great War.