With the exploits of the Medes, however, we are not here concerned. The two questions in connection with them which are of importance from our present point of view are— first, what was their language? second, what was their religion?

It has been hitherto assumed, in accordance with the most prevalent, and, in the opinion of the writer, the most The language of the Medes. probable view, that the Medes were an Íránian race speaking an Íránian language closely akin to Old Persian. This is the view taken, for instance, by Nöldeke, who, in concluding his account of the Medic Empire, says:* —

“Perhaps careful examinations of the neighbourhood of Hamadán, or excavations, may still some day bring to light other traces of that ancient time. It would be of the greatest value if inscriptions of the Medic kings should chance to be found; I should conjecture that these, both in language and script, would be quite similar to those of the Persian kings.”

Darmesteter, whose views will be discussed at greater length presently, goes further, and declares that the language of the Avesta, the so-called Zend language, is the language of Media, the Medic tongue.

“La conclusion qui s'impose,” says he,* after adducing evidence in favour of his view, “c'est que la tradition parsie et l'Avesta, confirmés par des témoignages étrangers, voient le centre et le berceau du Zoroastrisme, soit en Atropatene, soit a Raï, dans l'un et l'autre cas en Médie… Je crois que les droits de l'Atropatène sont mieux établis, et que c'est de là que le Zoroastrisme a pris sa course de l'Ouest à l'Est. En tout cas, le Zoroastrisme est une chose médique, et l'Avesta est l'œuvre des prêtres mèdes… Il suit … par le témoignage externe des classiques joint au témoignage intrinsèque des livres zends et de la tradition native, que l'Avesta est l'œuvre des Mages, que le zend est la langue de la Médie ancienne, et que l'on aurait le droit de remplacer le nom impropre de langue zende par le terme de langue médique.”

A totally different view, which ought not to pass unnoticed, is held by Oppert, and set forth at length in his work Le Oppert's view that the Medes were a Turanian race. Peuple et la Langue des Mèdes. The inscriptions of the Achæmenian kings, as is well known, are drawn up in three different languages, of which the first is Old Persian and the third Assyrian. As to the second, concerning the nature of which much doubt has prevailed, M. Oppert holds that it is Medic, and that it is not an Aryan but a Turanian tongue; which astonishing opinion he supports by many ingenious arguments. The very name of Media (Máda) he explains by a Sumerian word mada, meaning “country”; and the names of the Medic kings given by Ctesias he regards as the Aryan equivalents of the Aryanized Turanian names given by Herodotus and in the Old Persian inscriptions. Thus, for instance, in his view, the name of the first Medic king of Herodotus was compounded of daya (other) and ukku (law), the Aryanized or Persianized form of which was probably Dáhyuka, “le réunisseur des pays”; while the Persian translation of the same was the form given by Ctesias, <text in Greek script omitted> which “recalls to us the Persian Artáyu, from arta, ‘law,’ and áyu, ‘reuniting.’” Of the six tribes of the Medes mentioned by Herodotus (bk. i, ch. ci), Oppert admits that the names are Aryan; but he contends that in the case of two at least, the <text in Greek script omitted> and the <text in Greek script omitted>, we have to do with Aryan translations of the original names, which he believes to have been Túránian, and to have denoted respectively “autochthones” and “vivant dans les tentes.”

There are but very few scholars who are qualified to re­survey the ground traversed by M. Oppert and to form an Darmesteter's view. independent judgment of his results in matters of detail; but, as regards his general conclusions, we concur with Darmesteter in the summary state­ment of objections to M. Oppert's theory wherewith he closes his review of the book in question:* —

“Nous ne voyons donc pas de raison suffisante pour abandonner l'opinion traditionelle, que la langue des Mèdes était une langue aryenne, opinion qui a pour elle, en somme, le témoignage direct de Strabon, er le témoignage indirect d'Hérodote, sans parler des raisons très fortes qui font de la Médie le lieu d'origine du Zend Avesta et par suite la patrie du zend.”

In the absence of further discoveries, the theory that the Medes were an Íránian people speaking an Íránian language closely akin to Old Persian is the view which we must con­tinue to regard as most probable.

It has already been said that the Medic kings, unlike the Achæmenians, left no records of their achievements; while, as The Avesta. regards their language, some scholars, like Nöldeke, think that, though specimens of it may be brought to light by future discoveries, none are at present accessible; others, like Oppert, find such specimens in the cuneiform inscriptions of the second class; while others, like Darmesteter, believe that we possess in the ancient scriptures of the Zoroastrians, the Zend-Avesta, an ample specimen not only of the language, but also of the literature, of the Medes. That the language of the Avesta is an Íránian language, standing to Old Persian in the relation of sister, not of daughter or mother, is proved beyond all reasonable doubt. As to the part of Írán where it flourished, there is not, however, the same unanimity; for while Darmesteter, as we have seen, regards it as the language of Media, the opinion prevalent in Germany is that it was the language of Bactria, and it has even become fashionable to speak of it as “Old Bactrian” and “East Íránian.” Darmesteter, in his usual clear and concise way, sums up the arguments of the East Íránian or Bactrian theory before proceeding to refute them, as follows:* —

(1) Zend is not the language of Persia.

(2) It is in Bactria that, according to tradition, Zoroaster made his first important conquest, King Gushtásp.

(3) The geography of the Avesta only knows the east of Írán.

“The first fact,” he continues, “is correct, but purely negative; it excludes Persia [i.e., Persis proper] from the question, but leaves free all the rest of Írán.

“The second fact is correct, but only proves that Bactria plays a great part in the religious Epic of Zoroastrianism; the struggles maintained by the Íránians against the idolatrous Túránians, of which Bactria, by its geographical position, was the natural theatre, must necessarily have drawn the thoughts of the faithful to this part of Írán, where the worshippers of Ahura Mazda were at death-grips with the worshippers of the daêvas, and which formed the frontier-post of Ormazd against barbarous idolatry; it is even very probable that the legends concerning the conversion of Bactria and of King Gushtásp bequeath to us a historic recollection of the con­quests of Zoroastrianism in the East. Nowhere, however, is Bactria represented as the cradle of Zoroaster and Zoroastrianism; Pársí tradition is unanimous and consistent in placing this cradle, not in the East, in Bactria, but in the West, in Atropatene; and not only Pársí tradition, but the Avesta itself, for—

“The third fact adduced is incorrect: the Avesta knows the North and West of Írán as well as the East: the first chapter of the Vendidâd, which describes Írán as it was known to the authors of the Vendidâd, opens the enumeration of the Íránian regions by the Êrân-Vêj, washed by the Good Dâitya (I, 3); now the Êrân-Vêj is on the borders of Atropatene, and the Good Dâitya is the Araxes.* It is equally familiar with the North, for it cites Rhagæ, the <text in Greek script omitted> of the Greeks, the Ray of the moderns, in Media.”

One piece o?? philological evidence is adduced by Darme-steter in support of his opinion that the language of the Avesta is the language of the Medes. The modern Persian word for dog, sag, implies, says he,* the existence of an Old Persian form saka (not actually occurring in the meagre documents on which we depend for our direct knowledge of the ancient language of Párs). Herodotus, however, mentions (I, 110) that in the language of the Medes the dog was called <text in Greek script omitted>, which rather resembles the Avestic word span (San­skrit svan, Greek <text in Greek script omitted>). And it is curious that this word, in the form ispa, still exists* in some of the Persian dialects, such as those of Qohrúd (near Káshán) and Naṭanz. M. Clément Huart's develop­ment of Darme­steter's view. Huart, who has contributed to the Journal Asiatique* a number of very ingenious and interesting papers on various Persian dialects, such as those of Yazd, Síwand, and the curious Jáwidán-i-Kabír (the principal work of the heretical Ḥurúfí sect,* which arose in Persia in the fifteenth century of our era), has still further developed Darmesteter's views, and has endeavoured to show that several of the dialects spoken in remote and mountainous places in Persia (especially in the West, i.e., in Media) are descended from the language of the Avesta; and to these dialects he proposes to apply the term “Modern Medic,” or “Pehlevi-Musulman.”* He remarks that, amongst other differences, the root kar- underlies the whole verb which signifies “to do,” “to make,” in the Avestic language; while in Old Persian the aorist, or imperative, stem of this verb (as in Modern persian) is kun-; and again that the root sig­nifying “to speak,” “to say,” in Avestic is aoj-, vach-, while in Old Persian it is gaub-. Now while in Modern Persian (which, as we have seen, is the lineal descendant of Old Persian) the verbs signifying “to do,” “to say,” are kardan (imperative kun) and gujtan (imperative gú, gúy), in those dialects which he calls “Modern Medic” the stem kar- is preserved throughout (aorist karam instead of kunam, &c.), and words denoting “speech,” “to speak,” are derived from a root váj- or some similar basis corresponding to the Avestic aoj-, vach-. This test is employed by M. Huart in classifying a given dialect as “Medic” or “Persian.” According to this ingenious theory the language of the Avesta is still represented in Persia by a number of dialects, such as those used in the quatrains of Bábá Ṭáhir (beginning of the eleventh century), in the Jáwidán-i-Kabír (fifteenth century), and, at the present day, in the districts of Qohrúd and Síwand, and amongst the Zoroastrians of Yazd and Kirmán. It is also to be noted that the word for “I” in the Tálish dialect is, according to Berésine,* az, which appears to be a survival of the Avestic azem (Old Persian adam). It is to be expected that a fuller and more exhaustive study of the dialects still spoken in various parts of Persia (which, notwithstanding the rich materials collected, and in part published, by Zhukovski,* are still inadequately known to us) will throw more light on this question. Darmesteter, however, in another work (Chansons populaire des Afghans, pp. lxii-lxv), has endeavoured to show that the Pashtô or Pakhtô language of Afghanistan represents the chief surviving descendant of the old Avestic tongue, which theory seems to militate against the view set forth in his Études Iraniennes. It is possible, however, that the two are really compatible; that Zoroaster, of the Medic tribe of the Magians (Magush), brought his doctrine from Atropatene (Ázarbayján) in the extreme north-west of Írán to Bactria in the extreme north-east, where he achieved his first signal success by con-verting King Vishtáspa (Gushtásp); that the dialects of Atropatene and Bactria, and, indeed, of all North Írán, were very similar; and that in the Avesta, as suggested by De Harlez, the so-called Gáthá dialect represents the latter, and the ordinary Avestic of the Vendidâd the former. All this, however, is mere conjecture, which at best can only be regarded as a plausible hypothesis.