CHAPTER V. TRIBES ADJOINING SCYTHIA ACCORDING TO HERODOTUS AND ARISTEAS.

Eastward of the Neuri in the general description of Scythia⁰ and in the other passages where they are referred to, come the Androphagi. But in the account of the Neuri, c. 105, it is said that the latter, when invaded by snakes, migrated to the Budini, that is past Androphagi and Melanchlaeni.

name to mean Waterfolk, from vodd. Gelonus reminds him of the typical spread-out Slav settle­ment. His tradition is carried on by Niederle, Slav. Ant. 1. p. 266.

⁶ Her. iv. c. 102, 106.

⁷ This expedient of supposing doubled tribes is excused by many instances of tribes with similar names, especially in Eastern Europe, under condi­tions which make it easy for part of a nation to split off, e.g. Royal Scyths and colonist Scyths in Herodotus, three or four tribes called Huns, so too with Alans, Turks, Bolgars, Tartars, Kalmucks, Nogai, all of which have had subdivisions living at one time far apart from each other. This list might be almost indefinitely extended.

The Androphagi were probably Finns, and the most barbarous of them, as no trade route passed through their land. Theirs would be central Muscovy and southwards towards Chernigov. Hence, too, the most exag­gerated stories would be told of them. But we need not believe that they were cannibals any more than the Samoyeds, Finns also, whose name means the same. Tomaschek ingeniously suggests that the Amadoci of Pseudo- Hellanicus^[327] and of Ptolemy are the same as the Androphagi, amadaka, cf. Skr. amdd, eater of raw meat. He would propose to identify them with the Mordva of the present day, which is very possible, for there is no doubt that all the Finnish tribes now found on the middle Volga and on the Kama once lived far to the west or south.

If the Androphagi are Finns, Mordva, the Melanchlaeni are Finns also, Merja and Cheremis. The former were early absorbed by the advance of the Slavs, and the latter have been so strongly subjected to Turkish influence that all earlier traces have been wiped out. But archaeological evidence proves that some such tribe occupied the regiof) corresponding to that assigned by Herodotus to the Melanchlaeni about Riazan and Tambov^[328]. It may be a coincidence that the Cheremis wore black till a hundred years ago. Dark felt is the natural product of the coarse dark-woolled sheep of the country. So we need not see any connection with the SavSaparat of the Protogenes inscription (Ossete san black, daras garment) who were almost certainly a Sarmatian tribe. For the kind of name compare the Caucasian Melanchlaeni, who have tended to the confusion of later writers, and in modern times the Kara Kalpaks, White Russians, and such like.

Next to the Melanchlaeni and now above the Sarmatians, well to the east of Scythia, lived the Budini, fifteen days’ journey from the corner of the Maeotis. The Oarus seems to have flowed through their country, coming from that of the Thyssagetae. If then we measure fifteen days’ journey up the Don to the portage by Tsaritsyn and then up the Volga, we come to the lower part of the governments of Saratov and Samara, and not far to the north begins the forest region. The territory of the Budini probably included the lower courses of the Belaja, Vjatka and Kama. The inhabitants are most likely represented by the Permiaks, driven north and east by the spread of the Slavs and the irruptions of the Tartars.

Near the junction of the Kama and Volga there has always been an important trading post, Kazan since the coming of the Mongols, in early mediaeval times Bolgary. Gelonus seems to have been the first of the

v] Antlrophagi, Mclanc/ilaeni, Bicdini^ Ge Ionics 105

series. We have the name of another town among the Budini, Kapicrxo?^[329]. Tomaschek compares Permian karysok, little fortress. The wide commercial relations of this district are shewn by the wonderful silver plates found in the government of Perm, splendid specimens of Graeco-Roman, Syrian, Byzantine, Sassanian and even Indian work being dug up in these remote forests, as well as coins of Indo-Scythian kings^[330], evidence of connection with Central Asia. All these precious wares must have been paid for with furs. There may well have been a sufficiently lively trade to tempt the Greeks to establish a factory in the interior of the country, even as far from the coast as the land of the Budini^[331]. Herodotus probably exaggerated the number of the Greek population, as he has most clearly exaggerated the extent of the town of Gelonus. Three miles and a half square is an impossible size, three miles and a half about would be plenty for warehouses and temples and gardens and space for folding the local sheep of which Aristotle speaks. The establishment must have been like one of the forts in Canada, inhabited by a mixed population of traders and trappers, or the Ostrogi in Siberia, round which towns like Tomsk and Tobolsk have grown^[332]. The description of the Budini themselves tallies with that of the Permiaks, grey-eyed and reddish-haired: compare what Ibn Fadhlan says of the

Bashkirs, “Pediculos comeduut." The otters and beavers of Herodotus have become rarer with assiduous hunting, but they were common when the Russians first came, and found a home by the many rivers of the country⁵. His lake may be the marshes on the course of these, for instance about the lower Kama.

It is barely conceivable that the Neuri should have come so far for refuge as to the middle Volga, hence the probability of there having been other Budini near the Dnepr. These Darius may perhaps have reached ; Ptolemy’s Bodini seem the mere survival of an empty name.

Niederle⁶, while admitting that the Androphagi and Melanchlaeni are Finns, is inclined to think the Budini Slavonic. He regards them as stretching from the Dnepr to the Don behind the Androphagi, although Herodotus says distinctly that beyond these is a real desert and no men at all. Budini looks certainly very like a Slavonic tribe-name with the common suffix -in-, and there are plenty of Slavonic names from the root bud-. But they certainly stretched further east than Niederle allows, for they lived fifteen days up the Don above the Sauromatae. By bringing them west he puts Gelonus on the site of Kiev.

Europe in mediaeval times, v. Nuremberg Chronicle (1493), fol. CCLIII., “Sabatz in Hungaria.” The gorodishche or camp at Belsk (v. p. 147) excavated by Gorodtsov is 20 miles round, much larger than Gelonus.

^s Her. speaks of a marsh in which are taken tvvftpifs καί κάστορα καί άλλα θηρία ταραγωνο- πρόσωπα. The last I wrongly identified with the Tarandus or reindeer, v. sup. p. 5 and nn. 6, 7, but the marsh and Theophrastus l.c. rule this out. The ΐνύδρια usually translated otters are water­snakes, v. Pliny, NN. XXX. §21, xxxii. § 82, and the square faced beasts are the otters; a gloss to this effect has been misapplied.

⁶ op. cit. 1. p. 275.

M.

The late Professor I. N. Smirnov^[333] of Kazan, the chief authority on the Volga Finns, directly denies that the ancestors of the Cheremis and Mordva were the Melanchlaeni and Androphagi. But he does not advance any very valid objections, and admits a contact with Iranians which argues a seat further to the south. He denies any contact with Greeks such as we must suppose in the case of the Budini. Incidentally he describes many customs among the Finns that recall Scythian usages : among the Cheremis the sacrifice of a horse forty days after death and the stretching of its skin over the tomb: the soul does not really leave the body for forty days and even later comes back to it by a hole left for the purpose. On this fortieth day is the wake, at which the dead man assists, and is taken back to the grave on a cart with bells : among the Mordva again, after forty days there is a wake and a horse sacrifice and a washing of the funeral car. In both cases many things are put in the grave, or the dead will come and fetch away both things and people. This is all in favour of the existence of an Uralo-Altaic element among the Scyths, although there was a clear line of distinction drawn Between them and these Finns : for the Finns lived in the forest and the mixed multitude of Scythians in the steppe.

South of the eastern Budini were the Sauromatae, stretching east and north from a point three days’ journey to the east of the Tanais (which Herodotus takes to run southwards), and the same distance north of the corner of the Maeotis. Hippocrates says they are a special tribe of Scythian, and Herodotus, deriving them from a marriage of Amazons and Scyths, shews that they spoke a language akin to that spoken by the Scyths but gave their womenfolk more freedom^[334].

North-Easterly Trade Route.

■Herodotus derived his account of these nations, Agathyrsi, Neuri, Androphagi, Melanchlaeni, Budini and Sauromatae, from two sources and gives particulars of them in two places. In the one (cc. 100—109, Map iv. p. 27) he is keeping in view the story of Darius and his expedition, but these tribes, although set out according to the scheme of the square, are not wrested far from their places as given by the less detailed account which goes with the less schematic description of the lie of the land (cc. 16—26, Map v. p. 34). This he supplements with much information, partly due to Aristeas, as to tribes living in a north-easterly direction far into Central Asia (Map 1.). Due north of the European tribes Herodotus imagines a continuous desert, occasionally diversified with the lakes neces­sary for.the southward-flowing rivers : this desert is a real desert as opposed to the patches of thinly peopled land separating hostile tribes. Probably this real desert was actually uninhabited, as the forests of the far north were only peopled comparatively lately, when these very tribes were driven up by new comers from Asia, or the Lapps and Samoyeds crossed from the far ne.

me at Kazan, and especially to the translator, to whom I am indebted for my knowledge of Russian and for many favours, including the loan of this very book. Abercromby, op. cit., mostly follows Smirnov.

² Her. IV. 21, 110—117, also infra, p. 119, for their migration west of the Tanais.

Budini, Sauromatae, T/iyssagctac^ lyrcae 107

The land of the next tribe, Thyssagetae, is beyond a desert seven days’ journey across, lying to the N. or rather e. of the Budini*. From their country run the four rivers Lycus, Oarus, Tanais and Syrgis into the Maeotis. This last detail is not to be reconciled with geography (cf. p. 30). We can only think that it was a country with several rivers running sw., down which people got to the Maeotis across the Tsaritsyn portage. This would give us the western slope of the Ural from Ufa to Orenburg. Herodotus says nothing of the Urals. Their incline is so gentle that they do not strike a traveller as mountains. Here is a river, Chussovaja, which may have the same root as Thyssagetae. The termination of this latter form is Scythian or Sarmatian, cf. Tyragctae, Massagetae. In Ossetian, -gd­is an adjectival affix and -la the plural termination. Tomaschek identifies the Thyssagetae with the Voguls.

The trade route described by Herodotus passed far to the north and crossed the Urals, avoiding the barren Caspian steppe. Herodotus knew that hereabouts was no channel leading to the Northern Ocean, and in this he was in advance of the more scientific geographers down to Marinus of Tyre.

To the south lived tribes of more or less Iranian affinities, Sauromatae, later Aorsi and Alans, marching with the Finnish and Ugrian tribes above them and with the Caucasians to the south. They carried on a profitable trade between the mines of the Ural and Iran, and also between the Mediterranean world and the Far East. In the Chinese annals the Yen-ts'ai or Aorsi, afterwards called A-Ian-na, held the country from the Aral sea to the borders of Ta-Ts'in (Roman empire), and their traders even reached China.

With the next tribe, the lyrcae (iv. 22), we get beyond the stage for the wanderings of Darius. They are interesting for their name, which can hardly be other than the Sarmatian form of Jugra^{[335] [336]}, the word whence we have Hungarian. The ancestors of the Magyars were a tribe be­tween the Voguls and the Qstjaks, swept from their place by the Turkish invasions ancUnow a racial erratic block in the middle of the Slavs. Here we have the first notice of them^[337]. Their peculiar method of hunting, repre­sented on a gold plaque in the Hermitage^[338], required a country full of trees but not a thick forest: such would be the basins of the Tobol, the Ishim, and the Irtysh, just to the e. of the southern Ural and the land of the Thyssagetae^[339].

As neighbours of the lyrcae, Herodotus speaks of a tribe of Scyths that had separated from the Royal Scyths of the Euxine Steppes. Considering the ease with which a nomadic nation divides and sends off one part to a surprising distance (e.g. the Kalmucks, the majority of whom in the reign of Catherine II. of Russia left the lower Volga for the frontiers of China³), it is impossible to say that a part of the Royal Scyths could not have migrated

The name Turk had not yet come into existence, though it would be no proved anachronism to say that races kindred to the Turks had passed this way. “Turcae” in the MSS. of Mela, I. 116, and Pliny, NH. vi. 19, may well be due to intelligent copyists.

^G De Quincey’s account is mostly fancy, but vividly presents the possible circumstances of the great migration. Corrections are made in vol. VII. of Masson’s Edinburgh edition.

io8 'Tribes adjoining the Scythia of Herodotus [ch. north-eastwards. That there is a connection between inhabitants of these mutually remote regions is rendered probable by the similarity of many objects found here on the upper waters of the Jenisei and in the Scythian graves. Perhaps an easier way of supposing the conditions is to imagine that here again travellers found a subject population ruled over by a tribe with customs and language similar to those of the original royal caste of the Scyths. It is hard to imagine Iranians so far to the north beyond the utmost bounds of the Aryan world. If the Scyths were Ugrian rather than Turko-Tartar, this would be just the place from which they should come. The Scythian traders finding these Scyths far in upper Asia recalls how the mediaeval,. Magyar missionaries found again their kin the Voguls and Ostjaks.

yi rgippaei.

As far as these Scyths, says Herodotus, all the land is flat and deep-soiled ; henceforward it is stony and rugged. That is, we are coming to the outliers of the Altai mountains. On the upper Irtysh the steppe ceases about Bukhtarminsk. The trade route from the Ural came down from almost a north-westerly direction, and continuing the line we should be brought to Dzungaria and the country about Kuldzha well described as lying beneath lofty mountains, the Altai on one side and the Then Shan on the other. Here we meet with the Argippaei (c. 23), (the exact form of the name is uncertain : Argimpaei, Arimphaei, Orgiempaei, etc.). To the e. of them again, or rather to the se. following the same general line, come the Issedones^[340]. The position of the Issedones can be approximately fixed from Ptolemy’s account which has been well interpreted by Tomaschek^[341] as placing them in the Tarim basin. That is that the northern route followed by the informants of Herodotus, and a more direct way by which went Maes Titianus, the Syrian merchant, bring us to the same region.

In the Argippaei we have undoubtedly pure Mongols. Herodotus says of them that they are bald from their birth both men and women, have flat noses and large yez^eta, translated by Tomaschek cheek-bones, and speak a language of their own, but wear the dress of the Scythians. The baldness may well be a misunderstanding of the custom of shaving the head, or an exaggeration of the scantiness of hair which distinguishes the Mongolian race: the other details point clearly to Mongols and are borne out by what is told us of their food and manner of life^[342].

They live off a tree called Ponticum about the size of a fig tree, bearing a fruit like.a bean but with a stone. When this is ripe they rub it through a cloth and a thick black juice runs off from it. This juice is called Aschy. This they use as it is or mix it with milk, and. of the pulp of the fruit they make cakes and eat them. For they have not much cattle as their pastures are not excellent. This ponticum seems to

v] Eastern Scyths. Argippaci 109

be an Iranian word meaning the way-tree, “travellers’ joy” as it were: but “ aschy ” is Turkish and seems closest to dci, sour¹. It appears to be the Bird Cherry, Primus Partus, which is treated in exactly this way by the Bashkirs. But many other steppe berries are similarly used by various tribes.

The tree covered with felt in the winter is a picturesque account of the felt tent supported by a light and portable framework now universal among the nomads of Asia. It has entirely superseded the waggons in which the Scyths lived, being more roomy, more adaptable and in every way superior, except that it has to be taken up and down, and affords no shelter during the actual journey (v. supra p. 32 and f. 7).

The most remarkable point about the Argippaei is the respect in which they were held by their neighbours. Says Herodotus, “No man at all wrongs these men. For they are said to be sacred. Nor have they any weapon of war. And they both act as adjusters of differences among their neighbours, and if any man take refuge from pursuit with them he can be touched by no one.” Tomaschek supposes that these were the frontier officials of a well-organised Turkish kingdom, set to prevent the interruption of commerce by the quarrels of the various tribes upon its borders.

In general, however, the Greeks had a tendency to idealize the life of nomads. One might almost say they found in them the noble savage. Hence Homer speaks of the Mare-milkers as the most just of men², and Strabo (xi. viii. 7), speaking in particular of the Massagctae, but in general of all who live in Scythic wise, says, “ Such have a manner of life common to them all, which I have often spoken of, and their burials are much the same, and their customs and all their life together, independent but rude, wild and warlike, however as to contracts they arc straightforward and honest.” So the Chinese speak alternately of the treachery and honesty of their nomad neighbours.

Herodotus (iv. 24) says that all is perfectly clear and definite as far as the bald people, that Scyths and Greeks from the Pontic trading towns can tell about them ; further that these Scyths use seven interpreters to make their way through seven tongues. It is not quite clear how the number seven is made up. The tribes that may come in are Scyths, Sarmatae, Budini, Geloni, Thyssagctae, lyrcae, other Scyths, Argippaci and perhaps Issedones. In such a tale there is a great temptation to bring in as many tongues as possible, and the informants may well have reckoned in the Scyths themselves, or made Sarmatian into a separate language, or likewise Eastern Scythian, or counted in the Geloni, whatever their jargon may have been : in any case seven is a fair total, though five would probably have done.

Beyond the Argippaei (c. 25) to the north as it seems arc indeed great and high mountains, the main ranges of the Altai: the goat-footed men need not be snow-shoe men, as Tomaschek suggests, but any active mountaineers, and the folk who sleep six months in the year always mark the bounds of knowledge or rather inference towards the north.

Or aksi, Vdmbdry, op. cit. p. 98.

² II. XHi. 6.

Issedones.

To the East, or rather se. of the Argippaei, are the Issedones (c. 26)^[343], apparently Tibetan tribes in the Tarim and Bulunggir basin.

The customs of these people as related by Aristeas exactly recall those ascribed by mediaeval and modern travellers to the Tibetans. As Zenobius sums it up (v. 25) the Issedones eat their parents except their heads: their heads they cover with gold. Compare Rubruck translated by Hakluyt (p. 116):

“Next vnto them” (i.e. the men of Tangut) “are the people of Tebet men which were wont to eat the carkases of their deceased parents : that for pities sake they might make no other sepulchre for them but their owne bowels. Howbeit of late they have left off this custome, because that thereby they became abominable and odious vnto all other nations. Notwithstanding vnto this day they make fine cups of the skuls of their· parents, to the ende that when they drink out of them they may amidst all their Jollities and delights call their dead parents to remembrance. This was told mee by one that saw it. The sayd people of Tebet haue great plentie of golde in their land.” In the British Museum may be seen skull cups richly mounted such as are used in Tibet in the Lamaist ceremonies.

Further io'OK’jodrcc? 3e 6/zoiws at yvva'i.Kes tolclv dvZpdcnv. Not so much as it seems from their taking part in war and chase like the Sarmatian women, as from the importance naturally gained by the one woman of a polyandrous household. The Chinese even speak of states in this region in which the women held all the political authority.

If the testimony of Ptolemy according to all interpreters could not be adduced for putting the Issedones on the Tarim the positions of all the tribes along the trade route would lose a very important confirmation. The chief difficulty is that the Chinese describe wholesale changes of population as occurring between the times of Aristeas and of Ptolemy : the encroachments of the Hiung-nu (v. pp. 92 and 121) had in the second century b.c. driven the Ytie-chih from the Bulunggir basin into that of the Tarim. The Ytie-chih are said to have customs similar to those of the Hiung-nu, but polyandry is ascribed to them and they appear rather to have been nomad Tibetans, perhaps with Hunnish chiefs, at least they use the Turkish title jabgu. To the west of Lop-nor they found a town-dwelling population called T'u-huo-lo (Tochari)^[344]. Later we meet with both peoples in Trans-Oxiana and Bactria (hence the name Tokharistan) and they apparently leave the Tarim basin to the Hiung-nu^[345]. Had not the Ytie-chih been driven out of the country long before Ptolemy’s time their identification with his Issedones would be

³ They cannot have been cleared out com­pletely. We know that some, the Little Yue-chih, remained behind among the Tibetan K'iang. The inaccessible oases of the Tarim basin have har­boured the relics of many races. From his last journey Dr M. A. Stein brought back MSS. in twelve languages (Tinies, Mar. 8, 1909), but the Tibetan element seems the oldest at least along the South, having been present in Khotan before the historic invasion (Stein, Ancient Khotan, 1. p. 147).

Issedones, Tochari, IMassagetae

obvious : perhaps the name had clung to two settlements Issedon Scythica (Ak-su ?) and Issedon Se.rica (Lou-Ian near Lop-nor ?), reason enough for him to put the well known tribe on to his map. This is not on a par with his haphazard insertion of antiquated names towards the edges of Sarmatia : he had, as I shew below, a very good knowledge of the Tarim basin'.

So Ptolemy’s Issedones represent the Yue-chih in their second position on the Tarim, but Aristeas knew them on the Bulunggir and probably included the Tochari under them. So his Issedones might extend to the Pamir, where they would be opposite to the Massagetae just over the pass into the Jaxartes basin^{[346] [347].}

Massagetae.

Like tales are told of the Massagetae N. of the Oxus, of their way of eating their parents, not even having left them to die a natural death, and of their marriage customs^[348]. They are described as living opposite the Issedones^[349], that is, just across the mountains to the west of them, and are often coupled or even confounded with them. In iv. 13 Herodotus says, when speaking of the movement that drove the Scyths out of Asia, that according to Aristeas the Arimaspi attacked their neighbours the Issedones, and these drove out the Scyths : whereas in c. 11 he says that the Scyths were pressed by the Massagetae. The Massagetae are evidently a mixed collection of tribes without an ethnic unity, the variety of their customs and states of culture shews this, and Herodotus does not seem to suggest that they are all one people. They are generally reckoned to be Iranian. But it is probable that at any rate part of them were practically identical with the Issedones : that just as the Yue-chih were driven by pressure from the Huns over the mountains into Bactria, so before them another Tibetan tribe had trodden the same path under the same pressure and gained the country of the two rivers : perhaps this was the very movement of which Herodotus and Aristeas speak. Other Massagetae may well have been Iranian, or as some thought⁴, much the same as the Scythians; whereas the inhabitants of the islands of the Araxes (Oxus or Jaxartes, v. sup. p. 30) were aboriginals connected perhaps with the tribes of the Caucasus. The picture drawn of the nomad Massagetae seems very like that of Scythians in a rather ruder stage of development. The tale of Tomyris may bring to mind either the Tibetan gynaecocracy or that of the Sarmatians. Certainly it appears more closely linked with the latter. The name Massagetae seems to mean belonging to the great (horde), and probably just as all the tribes north of

³ Compare Her. I. 215, 216, ywalna pev yapiei fKaaros, TavrrpTi tirinoiva xpeatvTai. o yap ^KvOas "EXXi/vfr Troiteiv, ov Skv&ii eitriv oi Troicnvres aXXn MaacrayeTai · tt/s yap tirt0vpr]a’T] yvvaiKos Slaathe same as the Scyths, but the latter having attained to the idea of exclusive property in women who had been seized in war, had passed out of the stage of community of women.

⁴ Her. 1. 201.

the Pontus were for the Greeks more or less Scythians, all the tribes that were under the “great horde” were regarded by the Persians, from whom the Greeks mostly got their ideas of the peoples on the northern border of Iran, as all more or less Massagetae ; again it may have been the Scyths’ name for them.

Sacae.

For we must confess that no word like Massagetae occurs in the Old Persian inscriptions in which as we should expect from Herodotus (vii. 64) we find Saka. In the epitaph of Darius at Naksh-i-Rustam (a) we have Saka Tigrakhauda, Saka Humavarka, and Saka \f\yai\_y ta\radaraya {transmarinip. Oppert explains Tigrakhauda as “cunning with arrows.” It is usually taken to mean “with pointed caps,” and Humavarka has been compared with ’A/zvpytoi; the transmarine Sacae may be beyond

either the Aral or the Caspian or even, as F. W. Thomas^{[350] [351] suggests, Lake Hamun, as well as the Euxine, so that we are not much helped.}

On the rock of Bisutun^[352] Darius says himself (v. 22) “I went against the land of the Saka...Tigris...to the sea: I crossed it on a bridge, I slew the enemy, I seized...by name Sakunka...! seized also other rulers”; but the lacunae make it impossible to know to what expedition this refers. Saka are also mentioned as having revolted. At Persepolis (1. 18) Saka are named as bringing tribute. But which of these may be among the varied nations sculptured we cannot say. Those whose clothes have any resemblance to Scythic dress have been reproduced (p. 59, f. 12). Most interesting is the figure at Bisutun inscribed lyam Sakunka hya Saka] “this is Sakunka the Saka.” But of his national costume only the cyrbasia is left him.

Arimaspians and Hyperboreans.

As far as the Issedones reached there was a quite practicable trade route, and as it seems nearly allied Iranian tongues served as a medium of intercommunication beside the native idioms. As far as the Issedones it is quite possible that Aristeas of Proconnesus penetrated. From them he heard of other men living yet further east, but what he tells of these shews that we are coming to the lands where travellers’ tales flourish with most luxuriance. In the quotation from the Arimaspea preserved by Tzetzes, the Issedones say, “Above us^[353] to the north dwell men whose borders march with ours, many are they and mighty warriors indeed, rich in horses, wealthy in sheep, wealthy in cattle, shaggy of hair, sturdiest of all men ; and each has but one eye in his fair forehead—the Arimaspi.” Whatever the word

and translates, putting all into nominatives which do not seem to me to scan. Latyshev (Sc. et Cauc. 1. p. 322 = Tzetz. Chil. vii. 686) gives them thus:

Ισσηδο'ι χαίτρσιν άγαλλόμίνοι ταναρσι....

καί σφίας ανθρώπους elvai καθυπ(ρθ(ν όμαιμους προς Βορίω, πολλούς τι κα'ι ΐσθλούς κάρτα μαχητάς, άφνΑους ιπποισι, πολΰρρηνας, πολυβοΰτας.... Οφθαλμόν ϋ' tv ΐκαστος εχα χαρίιντι μίτώπω. \αιτησιν λασιοι πάντων στιβαρώτατοι άνδρών.

means, whether or no it be a folk-etymology, we cannot go behind the statement of Herodotus that the Scythians took it to mean one-eyed. The Chinese still say of the Khalkas, these people have but one eye, one hand¹, thus describing their awkwardness, and some such metaphor probably lies at the bottom of this talc. Beyond the Tarim basin to the north, we come precisely to the cradle of the Mongolian race. In this region the Chinese annalists of the Chou (b.c. 1155—255) and Han dynasties put the Hien-yun or Hiung-nu stretching from Shan-si across the Sha-mo far to the north of the T'ien Shan range. These are they whom we know in Europe as the Huns. Shorn of the poetic epithets, the description of Aristcas applies to them. They often joined into a well-organised state as often destroyed by the dissensions of the tribes. When united the)' controlled the commerce between China and the west and regulated it. The Bald-heads of Herodotus (iv. 23) would be their outpost to the west. True, Aristcas calls the Arimaspi yai- tt/itip Xctcnoi, but the warriors may well have been unkempt, while the custom officials would be shaved and smooth. Also in that western part in the gate of Dzungaria there would not be the abundance of flocks and herds that marked them on their native plains. Whether the Issedones received of them gold from the eastern Altai, or whether it did not rather come from the south from the mountains above India, and whether the griffins are not the ants or baibaks, that according to the story threw the gold out of their burrows, is more than can be said. Certainly the representations of Arimaspians and griffins in art belong to Western Asia. The griffins come from eastern stuffs (= cherub}, and their name is Semitic ; the Arimaspians arc dressed in barbarian costume, as conceived by the Greeks, on the model of the barbarians most familiar to them, Phrygians and Persians. Still the subject was felt to belong to Scythia, and was used to adorn goods destined for the Scythian market.

Beyond the griffins, says Aristeas², live the Hyperboreans, reaching down to the other sea. Herodotus doubts this, for he says he heard nothing about them from the Scyths³. The Hyperboreans are always the people beyond knowledge towards the north. They must always figure as the last term of any series that stretches in that direction. Still, as Tomaschek suggests,

¹ Cf. the Hou-yen-kuo, lit. Back-eye-people.

² Her. iv. 13, 32 ; Damastes ap. St. Byz. s.v.

³ By ignoring Ptolemy (v. p. 114, n. 3) F. W. Thomas (op. cit. p. 197) puts the Issedones in Farghana and the Arimaspi ( = Ariaspi) in Sistan, which hardly suits Her. and his KaruTrcpde. F. Westberg, Klio (Bd iv., 1904, pp. 182—192, “ Zur Topographic des Herodots”), by giving up the same point d'appui and restricting the area under consideration, has to use excessive ingenuity in fitting in the various tribes. He puts the Budini about Saratov on the steep or right bank of the Volga, and identifies them with the Burdas of Ibn Rusta; further he believes that Darius reached the Volga in this region. The desert above them is the high ground of the Zhiguldv Hills, and the Thyssagetae are on the Samara bend of the Volga and about the lower Kama with the lyrcae on the Belaja and in the southern Urals. The Bashkirs are the Argippaei, although they would appear to have reached their present position only in some

M.

later migration of Turko-Tartaric peoples. The names due to Aristeas he regards as mere alterna­tives of other tribal names known to Herodotus, so he identifies the Massagetae, whom he puts N. of the Jaxartes, with the Arimaspi, and opposite them the Issedones, whose women were so independent, with the Sarmatians. The Araxes of Herodotus I. 201 is for him the Jaxartes, but in c. 202, the Volga with its delta among the islands of which the fish­eaters live, and the Rhoxolani, whom we meet in later times, are *Araxalani, called after the river. Such a scheme seems to me to wrest the data given by Aristeas and Herodotus from their natural meaning, whereas something like Tomaschek’s view is far less arbitrary. Most original is d’Arbois de Jubainville (op. cit. I. p. 241 note); he supposes that the Arimaspi migrated from upper Asia to the Alps or Rhipaean mountains above Friuli. His object is to identify the Kelts with the Hyper­boreans.

15

some faint account of the civilised empire of China may have penetrated to Aristeas or his Issedon informants.

Aristeas also mentions the Rhipaean mountains, but again Herodotus does not believe in these. He is right in rejecting them to the north of the Euxine, but in upper Asia the difficulty is rather that among so many ranges we cannot tell which was intended by the name.

Always it has been at the apogee of the dominion of some Turko- Tartaric tribe that it has been possible for westerners to traverse central Asia. The voyage of Aristeas (c. 650 B.c.) comes at the time of the early nomad power which troubled the Chinese under the Chou dynasty. Those of Zemarchus and the Nestorian Alopen coincide with the greatest extent of the early empire of the western Turks which likewise, gave Hiian Tsang his opportunity to journey westwards¹. De Plano Carpini, Rubruck and Marco Polo were enabled to travel by the organization of the great Mongol Empire², and since its fall, till the other day, no European had followed"in all their footsteps, just as for seven hundred years no Greek followed Aristeas³.

¹ Chavannes, E., “ Documents stir les Tou-kiue (Tures) occidentaux recueillis et comments,” in Results oj the Orkhon Expedition, St I’., 1903.

² v. Heyd, Hist, du Commerce du Levant au Moyen Age, Paris-Leipzig, 1886, Vol. II. p. 215 sqq.

³ Ptolemy’s Serica. As it is physically pos­sible for Aristeas to have penetrated as far as the Tarim, the question whether Ptolemy’s Issedones can guide us in locating his, is a question of how much real knowledge of Central Asia Ptolemy shews, and requires a brief examination of his map of Serica &c. in the light of recent travel (Ptol. Geogr. I. xi. xii., VI. xiii.—xvi., cf. Maps I. and vi. ; see Yule, Cathay and the Way Thither, pp. xxxix. and c.xlvii.; Richthofen, China, 1. p. 477—500 and Map 8 ; Bunbury, Hist. Anc. Geogr., II. p. 529 sqq.; Tomaschek, op. cit. I. p. 736; Marquart, “Eran- sahr” (Abh. d. k. Ges. d. Wiss. zu Gottingen, Bd ill. No. 3, Berlin, 1901), p. 154, and lastly for one or two new points M. A. Stein, Ancient Khotan, I. p. 54). Richthofen seems most nearly right, Tomaschek gets everything too far to the E., Marquart hopes too much fora mechanical formula for reducing Pt.’s degrees to modern measurements. This is a fundamental mistake. In these regions all his knowledge is derived through Marinus of Tyre from Maes Titianus (60—80 A.D.) a Syrian merchant, who himself appears to have reached the “Stone Tower” and to have sent agents on to Sera Metropolis seven months’ journey: this dis­tance Marinus naively reckons at 36,200 stades, which Pt. is about right in halving, but this he does on general grounds, not on definite informa­tion. Hence we cannot take figures beyond the “Stone Tower” seriously. The more important is it that his map gives the general shape of the Tarim basin very fairly. The Imaus is clearly the Pamir, the only cross range in Central Asia, though of course it does not stretch indefinitely northwards. On it is set όρμητήριον των is την Σήραν Ιμπορίυο- μίνων, and Irkeshtam the Russian Custom-station commanding the passes both towards Farghana and towards the Alai plateau (Stein, p. 55) suits very well; the “Stone Tower” 5°W. must be on the Alai road. The Αΰζάκια ορη are the T‘ien Shan and the Άσμιραΐα to the S.E. of them the Kurruk Pagh : the Κάσια ορη are not about Kashgar but the W. K‘un-lun from which comes jade (Turk, kash); Thagurus Mons to the E. is Altyn Tagh or perhaps Nan Shan. Between the two mountain lines flows the Oechardes or Tarim with its im­portant source (Ak-su ?) in the T‘ien Shan, its sudden turn S. towards Lop-nor (>) ών Ιπ'ι τα Κάσια ορη έκτροπη, ν. S. Hedin’s Map in Through Asia II.), and its tributary from the K‘un-lun, the Charchan Darya ; the eastern part of it with its source in long. 174°, lat. 47⁰ 30', would perhaps be the Bulunggir which Stein says once joined the Tarim in Lop-nor (Geogr. Journ. XXX. 1907, p. 503 : Dr Hedin tells me he doubts this). Outside the Tarim basin the physical features are not so clear, but we may recognise the "Αννίβα ορη as the Altai, Enrodus and Ottorocorrhas* as the Himalaya and North Tibetan ranges, the Bautisus being the Upper Brahmaputra, but information as to this Southern side came from India, and Pt. not realising Tibet has made this river one with the Huang-ho crossed by the agents of Maes towards their journey’s end. The limit between Scythia extra Imaum and Serica represents if anything the extent of Chinese power in the 1st cent. A.D.: Αυζακία ττόλυ may be Kashgar and Issedon Scythica, Ak-su ; Issedon Serica, Lou- Ian by Lop-nor (Tomaschek brings it within the old western extension of the Chinese wall, Stein, l.c.); the Issedones between, the memory of the Yue-chih : it is tempting to see in Xavpava, Khotan and in 'Αχάσα χώρα the Khasas, confusion being produced by the combination of Indian and Serie information: the Thaguri though far to the E. may represent the Tochari (v. supr. p. 111, n. 1): Aspa- carae would be an Iranian term for nomads pro­bably Tibetan; Bautae, the Indian for Tibetans, cf. Bhotan, Sera Metropolis is more likely Ch‘ang- ’an the capital of the Elder Han near Si-’an-fu, than Lo-yang in Ho-nan Pt.’s Sinae Metropolis (long. 180"). The Annibi, Garinaei (? Mountaineers) and Rhabbanae would be Huns and perhaps Sien- pi ; the Siftryrs·, Kao-ch‘e. It is noticeable that fancy names like Abii and Anthropophagi are con­fined to the N. border of the map, so Issedon is not of that class.

* Please correct Otto(ro)corrhas, etc. on Map VI.