The most noticeable examples of the accuracy of his personal observation, and of the diligence and care of his inquiries from persons who were present at events, is afforded by the correctness with which he describes the topography of Thermopylæ and Platæa, and the close agreement of the incidents of those battles as recorded by him with the nature and accidents of the ground at the present day. There is no single statement of pure fact in his accounts of those battles upon which suspicion is thrown by a minute examination of the areas within which they were fought. The majority of them are positively capable of proof by the results of such an examination. The manifest errors in these two accounts do not lie in statement of fact, but in statement of motive, and are fully explicable by the circumstances under which the story was told. In neither case does he produce any evidence that he was able to get information as to the motives which influenced those in command. The story of the fighting in both cases is a soldier’s story. The details of the first two days of the defence of Thermopylæ were derived from some one who was present at, or a spectator of, the fight, of which there were many survivors. Those of the final scene are of the same character, though in that case it is more difficult to conjecture who was the informant. AUTOPSY AND TOPOGRAPHY. From certain indications in the story, it seems that parts of it at any rate may have been derived from some one who viewed it from the Persian side, possibly some Malian who had been temporarily impressed as a servant, and with whom Herodotus conversed when he visited the spot in later years. Such are the details (vii. 208) with regard to some of the Lacedæmonians being hidden from view of the Persian scout by the interposition of the wall, and some of the information given with reference to Epialtes (vii. 213). It is quite beyond question that Herodotus had traversed the ground. His topographical information can be followed in detail by any one who visits the region at the present day.

The only uncertainty concerns the exact sites of one or two places which he mentions outside the area of fighting, such as the Amphiktyonic temple; but even these are identifiable with high probability; and it is the use of the spade, not Herodotus’ information, which is wanting.

The rest of his information with regard to Thermopylæ, with the sole exception of the copies of the inscriptions on the monuments, is traditional in origin, emanating from Sparta.[232]

Artemisium he may have seen in the distance in the course of a land journey from the north. The knowledge which he displays of Halos, in his account of Xerxes’ visit to the place, renders it probable that he had traversed the route which leads round the east end of Othrys on to the Malian plain, and in going that way he would pass in full view of Aphetæ and Artemisium. His topographical description of these places is, in so far as it is called for, accurate enough. That is, however, the only portion of the Artemisium story which can be set down to aught of the nature of personal examination. Nor, indeed, can any part of that story be safely attributed to the evidence of an eye-witness of events. The list of the contingents of the Greek fleet may be taken from official records of some sort; the rest of the story seems to be drawn from tradition.

The tale of Salamis is, perhaps, the most extraordinary in the whole of his history. There can be little doubt that when he wrote it he had in his possession notes of information given him by some one who was present at the battle. Probably he never compiled its history of the battle until some years after he had taken the notes, at a time when his informant was not within his reach, possibly dead and gone. He must have known the topography of the strait; he could hardly have lived in Athens without doing so.

How did he come to make the serious error which he undoubtedly did make? There may be differences of opinion as to the nature of the error, still more as to its cause; but as to the existence of it the topographical evidence, as well as the contemporary account of the eye-witness Æschylus, leave no reasonable doubt. The mistake does not appear, in so far as can be judged, to have existed in his original information. His account, despite the confusion inevitable to the nature of his mistake, points to its being founded on evidence which may have been in all essential particulars in accord with that of Æschylus and Diodorus. The matter has already been discussed at length in reference to the battle itself. Suffice it now to say that the most probable explanation of it is that, using his notes of information some time after they had been given him, he mistimed a movement mentioned in them, and by so doing threw his narrative out of gear.

In the Salamis story, as elsewhere, the traditional element is present in various forms. As in the case of Artemisium, the list of the ships may be founded upon an official basis.

The story of Platæa is, perhaps, the most remarkable that he wrote. How he ever arrived at the truth with regard to that scene of confused fighting must excite the wonder of every student of history. FACTS AND JUDGMENTS. It is plain that he was peculiarly well satisfied with his evidence, for he was not in the habit of treating at length incidents whose description rested on testimony with which he was dissatisfied, however important these incidents might be. His account of Marathon affords proof of that; and there other ominous silences in his narrative which can only have been due to lack of reliable information. His topographical evidence on Platæa is undoubtedly due to autopsy. Not merely are his direct statements on the subject marvellously correct; but the whole of the topography implied in his account of the many incidents of the battle is marvellously correct also. His information was manifestly derived from one who had fought at Platæa; and the character of the story is such as to render it possible that he not merely visited the scene of it himself, but visited in company with one who had fought there. His account exemplifies in a peculiar way both his excellences and his limitations as a military historian. His own diligent inquiry led him to make the best use possible of the evidence at his command; but that evidence was not obtained from one who was conversant with the plans of those who directed the operations, nor had the historian himself such knowledge and experience as would enable him to form sound inductions as to the nature of those plans from information of the kind which was at his disposal.

The traditional element in his account of Platæa is probably far less in proportion than that in any one of the longer sections of his history. Of official record only the list of the Greek contingents shows probable traces.

Such are the major parts of his history in which the evidence of autopsy on his part, or on the part of his sources of information, is most strikingly displayed.