from its mouth, though below the cataracts there can never have been any obstacle to the navigation, which is indeed particularly easy. (See Clarke's Travels, l.c.)

NOTE E, p. 186.

RIVERS OF SCYTHIA.

Professor Rawlinson justly observes that "there is the utmost uncertainty with respect to all identifications east of the isthmus of Perekop," the Tanaïs alone forming the exception. It is simply impossible to accept the statements of Herodotus as they stand, and even the favourite resource of modern commentators—to suppose that great physical changes may have taken place in the countries in question-will do little to remove the difficulty. The main point of all is the Gerrhus, which is clearly described as branching off from the Borysthenes, at the place of the same name, which was the farthest point of that river that was known to the Greeks (iv. 56). An ancient geographer would see no improbability in this, as such bifurcations were supposed to exist in other cases, as that of the Ister, which was believed by geographers much more advanced than Herodotus to send off an arm to the head of the Adriatic, while the main stream flowed into the Euxine. But there is in fact no such case known in physical geography.

The supposition of Professor Malden (cited by Sir R. Murchison, Russia and the Ural Mountains, vol. i. p. 574) that the Borysthenes may have originally formed a delta, and reached the sea by two separate mouths, is not destitute of plausibility; but it would offer a very inadequate solution of the difficulty. For the formation of such a delta is only possible at a short distance from the mouth of a great river, or where (as in the case of the Euphrates and Ganges) it flows for a long distance through marshy and alluvial lands. But the region where Herodotus places the river in question is the steppe country, which is throughout considerably elevated above the beds of the streams that traverse it. The formation of a gigantic delta in such a country may be safely pronounced to be physically impossible. And whatever may be thought of the distance assigned by Herodotus from Gerrhus to the sea (see Note D), the bifurcation must have taken place, according to his view,

a long way up the course of the Borysthenes, as he supposes the two rivers Panticapes and Hypacyris, both of them considerable streams (Tотаμοì оvνоμаσToí, c. 47 and 58), and taking their rise in two separate lakes, to have their sources and their whole course between the other two rivers, or rather arms of the same river! And he tells us, moreover, that the country of the agricultural Scythians (the Georgi), which was situated between the Panticapes and the Borysthenes, was three days' journey in width, and extended up the Borysthenes for eleven days' voyage (iv. 18), while that of the nomad Scythians extended fourteen days' journey between the Panticapes and the Gerrhus (Ib. 19).

Even if we abandon the attempt to explain or reconcile these strange statements concerning the course and connection of the three rivers, the difficulty remains that nothing corresponding to them can be found between the prescribed limits. No streams of any considerable importance are found between the Dnieper and the Don (except the Donetz, a tributary of the latter); and those which exist all flow into the Sea of Azov, not into the Euxine.

Nor do the statements of later writers throw any light on the subject. Both Pliny and Mela mention the name of the Panticapes, but their accounts of these Scythian rivers are a mass of confusion, and Dionysius, who describes it as flowing from the Rhipæan Mountains (Periegesis, v. 315), is in direct contradiction with Herodotus.

NOTE F, p. 187.

THE RIVER OARUS.

The identification of the Oarus of Herodotus with the Volga was adopted by Major Rennell (Geogr. of Herodotus, p. 90, 4to. ed.), and has been accepted by most recent commentators. But it certainly rests on no adequate authority. If, indeed, the details of the expedition of Darius against the Scythians could be regarded as trustworthy, the supposition that the river Oarus, which was the limit of his progress towards the east, was the same with the Volga, would not be devoid of plausibility; but, as we shall hereafter see, those details are so clearly unworthy of credit, that no dependence can be placed upon this argument, and there is really no other. The supposed resemblance of the name Oarus to the Rha of Ptolemy, which is certainly the Volga, is so slight as to have no weight at

all. On the other hand, in the passage under discussion in the text he enumerates the rivers in an order which would seem to place the Oarus to the west of the Tanaïs; while in the account of the expedition he clearly represents Darius as crossing the Tanaïs, and advancing eastward to the Oarus. But he there does not mention the Lycus at all, though on this supposition Darius must have crossed it before coming to the Tanaïs, and we should thus have a fourth river to account for between the Tanaïs and the Borysthenes, without anything really corresponding to it.

This discrepancy in regard to the rivers would tend to show that Herodotus derived his account of the expedition of Darius from a different quarter from that which furnished him with the account previously given of Scythia and the adjoining countries, notwithstanding the close agreement already pointed out in the arrangement of the surrounding nations.

NOTE G, p. 190.

LIMITS OF SCYTHIA.

The limits here assigned are materially less than those adopted by Mr. Rawlinson, who appears to me to extend the Scythia of Herodotus much too far to the north. He seems to have been in part led to this conclusion by assuming that it comprised the "two great basins of the Don and Dnieper," as well as the "two minor basins" of the Dniester and Boug (Herodotus, vol. iii. p. 210). But we know from Herodotus himself that both the two great rivers had their sources far beyond the Scythian territory; the one rising in the land of the Thyssagetæ, separated by a broad desert from that of the Budini, who themselves lay to the north-east of Scythia proper; while the sources of the Borysthenes were unknown to him; but they were clearly situated beyond the farthest limits of Scythia, with the whole of which he considered himself as well acquainted.

On the other hand, M. Neumann, whose general views on the subject appear to me to be sound and judicious, carries them, I think, to an extreme, and is disposed to restrict the Scythians of Herodotus within narrower limits than is reasonable or necessary. Whatever value we may attach to his statement of their extending 4000 stadia, or 20 days' journey, inland, it is certain that he

regarded the Scythians as extending nearly, if not quite, to the sources of the Hypanis and the Tyras, as well as up the Borysthenes as far as Gerrhus. The hypothesis of M. Neumann, who would bring down Gerrhus below the cataracts of the Borysthenes, within a few days' journey of the sea, appears to me utterly at variance with the conception of Herodotus, who distinctly tells us (iv. 71) that the Gerrhi were the remotest tribe that was subject to the royal Scythians.

NOTE H, p. 195.

ETHNOGRAPHICAL RELATIONS OF THE SCYTHIANS OF

HERODOTUS.

There are few questions in ethnography that have been the subject of more discussion and controversy in modern times than the origin and ethnical affinities of the people described by Herodotus under the name of Scythians. The prevailing opinion is that they were a Mongolian race, like the Kalmucks in modern days; and this view has been adopted by Niebuhr (Kleine Schriften, p. 362; Vorträge über alte Geschichte, vol. i. p. 179); by Schafarik (Slavische Alterthümer, vol. i. p. 279), and by Neumann (Die Hellenen im Skythenlande, pp. 198, 199); as well as by our own historians, Thirlwall and Grote. On the other hand, several eminent philologers have contended that they were a people of Aryan or IndoEuropean race. Dr. Donaldson (Varronianus, 2nd edit. pp. 40-45) attempts to prove that they were a Slavonian race, like their neighbours the Sauromata; and Jacob Grimm (Gesch. der Deutschen Sprache, vol. i. p. 219) maintains that there is sufficient evidence to assign them to the Indo-European family, without venturing to determine the particular branch to which they belong. The same hypothesis is adopted by Alexander Humboldt (Kosmos, vol. i. p. 491); Professor Rawlinson (Herodotus, vol. iii. pp. 192-205); and by Zeuss (Die Deutschen und die Nachbarstämme, pp. 285–299). It may well be doubted whether we possess the means of arriving at a satisfactory solution of the question. On the one hand the elaborate account of the manners and customs, as well as the religious rites, of the Scythians, transmitted to us by Herodotus, presents so many points in common with those of existing Mongolian races, that it appears at first sight to be decisive of the

:

subject and the description of their physical characters by Hippocrates, who wrote but little after Herodotus, and undoubtedly refers to the same people (De Aere, Aqua, et Locis, c. 6, p. 558), tends to confirm this conclusion. But it must be admitted that there is so strong a general resemblance in the habits and manners of all the nomad races that have inhabited in successive ages the vast plains of Asia and Eastern Europe as to detract materially from the force of this argument. (See the judicious remarks on this subject of Humboldt, Kosmos, vol. i. p. 492.) On the other hand the linguistic grounds, which are principally appealed to by the advocates of the other theory, are scarcely sufficient to carry conviction to a mind not predisposed in their favour. The few words of the Scythian language quoted by Herodotus-who was himself wholly unacquainted with it-would naturally be liable to much distortion, and the same thing would apply to their proper names, which we possess only in the form into which they were altered by the Greeks. Philological conclusions based upon such slender materials are very far from possessing the conclusive authority which they justly claim when they rest upon a sufficient knowledge of the language.

The Scythians appear to have continued to occupy the regions north of the Euxine for some centuries after the time of Herodotus, but they gradually gave way before the advancing tide of the Sarmatian or Slavonian races from the east. Whether they were driven westward, or were gradually absorbed by the successive waves of nomad population that swept over their country, we have no means of determining; but at a later period they disappear both from history and geography, and Pliny's statement that in his time the (European) Scythians had become merged in the Sarmatians and Germans is probably well founded. (Scytharum nomen usquequaque transiit in Sarmatas atque Germanos, H.N. iv. 12, § 81.)

It is remarkable that the Alani, who were found in the fourth century after the Christian era in possession of the same tract, and whose manners, as described by Ammianus Marcellinus (xxxi. 2), present much resemblance to those of the Scythians of Herodotus, disappear in like manner from history, and their ethnical relations are almost equally uncertain.