will have developed the whole alphabet with an accuracy not to be attained except through a previous knowledge of the ancient languages of Parthia and Ariana.

The only types of the Pehlví character, which which we can institute a comparison of the above alphabet, are those derived from the imperfectly-deciphered coins of the Sassanian dynasty of Persian monarchs; and the inscriptions on the sculptures at Naksh-i Rustam, Naksh-i Rajab, and the Ták-i bostán. These are attributed to the same period, on the certain authority of the names of Bábak, his son Ardashir, and grandson Shápúr, found not only in the Pehlví, but also in the Greek version, which fortunately accompanies some of the inscriptions. The Baron de Sacy, to whose Mémoires sur les divers Antiquités de la Perse,' the learned world was indebted for the restoration of these valuable monuments of antiquity, was only able to deal with one form of the Pehlví, namely, that situated below the Greek (see Ker Porter, ii., 552): for the inscriptions are generally trilingual; the version above the Greek being more rude than the other, and having a striking resemblance to the Chaldaic. Ker Porter transcribes one or two portions of the upper inscriptions in Hebrew; and informs us that De Sacy always found this character had the same meaning as the Hebrew, when transcribed letter for letter. This author has given in pl. xv., vol. i. of his 'Travels,' a fine fac-simile in the two languages of the Naksh-i Rustam text, which had not yet been deciphered at the time of the publication of his work. A considerable portion of the members of our alphabet occur precisely in the right hand version of this transcript; such as 9, 4, 7, 1, 7, t, u, S, &, etc. but for want of a perfect alphabet, or of a Roman version of the inscription, no comparison can yet be made. The learned Lichtenstein, in his dissertation on the arrow-headed character, has furnished a plate of all the varieties of Pehlvi and Zend, as known in his time, from the travels of Niebuhr, etc. By way of exhibiting the analogy which exists between these and our new character, I have carefully set them in comparison, in pl. xi., taking Lichtenstein's imperfect alphabet of what he designates the ArabicoPersic Zend, as the only available one of this type. The Pehlvi inscription alphabet I have taken from Ker Porter's fac-similes; and the Pehlví of coins, from plates of coins in Marsden, Ker Porter, Hyde, etc., and from actual coins; but in most of the latter that I have seen, the letters are so very indistinctly formed, that it is quite impossible to read them; and, indeed, most of the attempts hitherto made have failed to pass the common titles: the names are very obscure. A reflection here forces itself that, if the coins of the Sassanian dynasty were so illegible, we need not be surprised at equal or greater difficulties attending those of the Bactrian princes.

In the sixth column I have inserted, at random, such of the letters on the cylinders, as approach in appearance to the coin types. No reliance, however, must be placed on this allocation, until a reading has been effected of some portion. It is only intended to show that the characters of the cylinders and coins are identical in their nature.

In the last column I have added the Zend alphabet, as restored with so much ability by M. Burnouf. It has a few points of accordance with the Pehlví; but the genius of it follows rather the Sanskrit type; and the constant expression of the

1 [Jour. As. Soc. Bengal,' iv., 1835. I have superseded this plate by one more suitable to the present stage of our knowledge of the subject. This engraving, as I have already intimated, will be reserved for the illustration of Art. XXI. The plate numbered xii. is, however, retained to mark its due position in Prinsep's fi...]

vowels, long and short, distinguishes it essentially from the alphabets of Semitic origin.1

Having thus completed our survey of the characters found on the Bactrian coins, and on the curious inscriptions extracted from the Topes (in which latter, however, we must expect to find such deviations from caligraphy as a written text naturally exhibits), let us now apply our uncertain knowledge, with circumspection, to the various names and titles on the coins themselves, and see how they may be read in Roman characters.

Plate xii. contains them all arranged,-first, according to the full inscriptions; secondly, with the names and titles separated. From what has been said above, I would venture thus to express the names of the Greek sovereigns in Roman letters. Apollodotou, Apaladado; Antilakidou, Atikalikado; Antimachou, Atimacho; Azou, Ajo or Ayo; Azilisou, Ajiliso; Eucratidou, Eukratido; Ermaiou, Ermayo; Menandrou, Minano or Midano; Philoxenou, Palatino or Palakino; Lysiou, Lisato, or Litato? Nonou, Ulalido? Unadpherrou, Fareto nanado? 2

It must be confessed that many of these are highly unsatisfactory, especially the last three. The name of Kadphises is omitted, as being still more indistinct.

Turning now to the titles and epithets, it does not seem difficult to recognise the same appellation for 'king,' and 'king of kings,' as is read on the sculptured inscriptions at Naksh-i Rustam, and on the Sassanian coins;-malako and malakao-malako (for malakan-malaka). When another epithet is introduced, such as the great king of kings,' it is found interposed between the words malakao and malako.

,מלכארברבא מלכא דאתור ,The same form of expression exists in the Hebrew

‘rex maximus rex Assyriæ. Every one will remark the close resemblance of this expression with our text; as well as of the Pehlvi title of the Persian sculptures-with PP; the terminations only being different, as might be expected in a different dialect. But, if the language of our coins be Zend, the word melek, 'king,' should not be expected in it; especially when we afterwards find it replaced by rao and rao nana rao, on the Kanerkos coins. It was this circumstance that led me to imagine the reading might be mahardo; but the combination maharáo-maharó is inadmissible, and overthrows the conjecture.

Pass we now to the next title of most common occurrence, P77, or 17 ΣOTHPOX, 'the Saviour.' By our system, this must be rendered either rakako, radako, or radado. Now the first of these three forms is precisely what might be expected to be the Zend reading of the Sanskrit word, rakshaka, ‘saviour,' and that alone is a strong argument in favour of its adoption as the true reading of the term.

The title, METAAOT BAZIAENE, first, we are told, assumed by Eucratides, belongs to so many of his successors, that we have no difficulty in finding the exact version of the term in the Pehlví. There are, however, decidedly two readings of it; one, P, the other Py, with the omission of the duplicated letter in the centre. The obvious rendering of these two expressions would be kák-kdo and kákdo. But I find in M. Burnouf's Commentaire,' that the Zend word for 'great' is maz, from the Sanskrit mahá, to which our term has no resemblance whatever. It has most simi

1 [In lieu of complicating the plate with this alphabet, I propose to insert hereafter (Art. XXI.) a type series of the Zend and such other Oriental Alphabets as may be calculated to throw light upon the special question of Bactrian palæography.] [As these have been, in many instances, rectified by Prinsep himself, I refer the reader to his latest paper on the subject: Art. XXI.]

ΒΑΣΙΛΕΩΣ ΝΙΚΗΦΟΡΟΥ ΑΝΤΙΜΑΧΟΥ PSump γυ
Ο ΒΑΣΙΛΕΩΣ
ΒΑΣΙΛΕΩΣ ΝΙΚΗΦΟΡΟΥ ΑΝΤΙΑΛΚΙΔΟΥ

ΜΕΓΑΛΟΥ ΟΝΩΝ ΟΥ

Η ΒΑΣΙΛΕΩΣ ΑΝΙΚΗΤΟΥ ΛΥΣΙΟΥ

larity to the Tartar appellation khákán, common to monarchs of Bactria in later days; and it may not be unreasonable to assume this reading, until one more plausible is discovered. At first I was inclined to read it ra-rao, in conjuction with the preceding word mahardo, on the strength of the expression before alluded to, of rao nano rao, on the Kanerkos coins; but in no example have I found the first letter replaced by. The collocation of the letters is, again, exactly similar to those of the Hebrew 7 rabreba, maximus'; but for this reading we must suppose y to be a b, which is contradicted by all other examples.

For ANIKHTOT, the unconquered, we find the terms apatilo, and apatilako, and for NIKH OPOT, 'the conquest-bearing,' ajaluko or ajalado of neither of these can I attempt a solution, and the examples being few, we cannot be very certain of their correctness

The inscription cut on the silver disc found in the casket of the Manikyala Tope, [vi.] (fig. 26,) may be read Pich Prub, famaro kanadako; the second word, without any very great straining, might be conceived to be the native mode of writing Kanerko; and if this interpretation be allowed, we may indeed look upon this tope as the monument of that monarch.

The writing on the brass cylinder itself (fig. 20 b of the same plate), which was, from my ignorance, inverted in the engraving, seems to consist of Bactro-Pehlví characters, which, rendered in Roman letters, would be kad.. malapo, far kamana papako, the purport of which I must leave uninterpreted: nor will I endeavour to forestal the ingenuity of others by any crude attempt to convert into Roman letters the longer inscriptions given by Mr. Masson, from the Jalálábád cylinder [vi.], and by M. Court, from the stone slab of another Manikyála tope (Jour. As. Soc. Beng., vol. iii., pl. xxxiii.) I have already remarked that this latter inscription contains, very legibly, in the second line, the word malakao, identical with the royal designation so common upon the coins.

It is now time to turn our attention to the coins themselves, whence our data for the construction of the Bactro-Pehlví alphabet have been derived. Of these I need do little more than furnish a few notes of reference to the accompanying six plates, in which I have brought down the series of selected specimens from Euthydemus to Kadaphes Choranos, a name so nearly allied to Kadphises, that the latter may be looked upon as its patronymic; while the title that follows it (choranos) coincides so closely with what has been already described as existing on the rao nano rao group (p. 131), that it would seem to form the link of connection between them and the coins which bear Pehlví legends on the reverse.

COINS WITH GREEK INSCRIPTIONS ONLY.

With Euthydemus of Magnesia, who conquered Theodotus II., B.C. 220, commences our present series: of his coinage I now possess a medal in silver, procured by Mohan Lál, for Dr. Gerard, near Kábul. It is superior in execution to the fine coin taken home by Lieut. Burnes. The exterior surface is of a dark-grey, like that of chloride of silver.

EUTHYDEMUS.
(pl. xiii.)

Fig. 1. Silver tetradrachma, weight 240 grs.
OBVERSE.-Head of the king in high relief.

REVERSE.-Hercules with his club, seated on clouds; inscription, BAXIAENE

ΕΥΘΥΔΗΜΟΥ.