much to excite interest in the broken sections we are already in possession of, that I confidently make the appeal to those who may happen to be in a position to improve our existing copies by means of photography, impressions, rubbings, new hand-tracings, or, better still, by intelligent transcripts in modern Pehlvi-for aid in the cause, towards which the portions of the text, now printed, will contribute something in the way of a first proof, and for the encouragement otherwise of future Palæographers, we may hope that, under a closer examination, the duplicate legends may aid each other both in defective passages and in the correction of the present disjointed order of sequence: while, as the first investigation was necessarily hasty, new discoveries of materials may happily reward more deliberate explorers, even as we can now appeal to the immense advance upon the imperfect transcripts of Niebuhr and Morier, achieved by the less hurried and amplified facsimiles of M.M. Flandin and Coste.1

In order to bring the entire subject under one view, I have collected together all the fragmentary inscriptions of the Sassanidæ at present known, commencing with those interpreted by De Sacy, which I simply reproduce in their corresponding literal equivalents in modern Hebrew and Persian type. The same course has been pursued with the highly interesting bilingual inscription of Sapor, from Hájíábád. Sir H. Rawlinson's unpublished copies of the Páï Kúlí legends, as well as his improved transcripts of the Ták-i-Bustán epigraphs have, however, been more exactly imitated in modern Pehlvi type, which has been made so far competent to resume its primitive duty by the introduction of three letters of the earlier alphabet, which have been lost in the degraded writing of the extant MSS., and finally a similar plan has been followed in the representation of the legible portions of two long and, for the present, most tantalizing inscriptions of Sapor: artists' designs of which have

1 Ker Porter remarks (i. p. 574), M. de Sacy "has followed Niebuhr's copy, which, strange to say, having been made so many years anterior to mine, exhibits an inscription much more defaced than I found it. This may be seen by comparing the large letters in my copy on the drawing with the large letters in M. de Sacy's Greek transcript." [Mem. sur div Ant. p. 31].

been given in Flandin's great work,1 though I am not aware that any attempt has hitherto been made to decipher or explain these singularly comprehensive documents. I am indebted to the same publication for the unique inscription of Narses, at Sháhpúr, which, together with the legends from the Royal signets of Varahrán Kirmán Sháh have equally been admitted to the honours of the adapted semblance of their contemporary Pehlvi.

None of the original drawings or published engravings of the more important inscriptions are sufficiently exact or continuously complete to recommend them for imitation in facsimile engravings, and even the plaster-casts from Hájíábád, however well they reproduce portions of the associate inscriptions, as exhibited in the Photograph, would not, in their present state, suffice to form an unbroken or perfect copy. The expedient has therefore been again adopted of recognizing these absolute impressions from the sculptured rock as a basis for the construction of standard alphabets of either class. In each case, the best examples of the normal character have been selected from the often-varying outlines of the same letter as fashioned by the local mason, and regard has always been paid to the corresponding outline of the given letter in other monuments of the period, whether lapidary, numismatic, or sigillary. The result has been embodied in the double column of alphabets engraved on wood, arranged with the ordinary type in the accompanying table; and, as in the absence of all other positive examples of lapidary writing, these letters have to play a conspicuous part as representative types of their several palæographic systems, no effort, short of cutting the individual letters, has been spared on my part to secure a true and effective rendering of the special characteristics of each symbol.

The primary derivation of these alphabets may obviously be traced to Phonico-Babylonian teachings. Specimens of that form of writing occur, so to say, in situ, as early as the time

1 Voyage en Perse, M. M. Eugène Flandin et Paul Coste, entrepris par ordre de M. le Ministre des Affaires Etrangères. D'après les instructions dressées par l'Institut. Paris, 1851. 6 vols. folio, plates, etc., and 2 vols. 8vo. text.

of Sargon, B.C. 721, when the individual characters present themselves in a fixed and cultivated form, far removed from the early stages of crude invention, an indication that, apart from the almost simultaneously established geographical range of cognate letters, would claim for them an extended anterior currency, which it would be as difficult to limit as to define; my own impressions have always leant towards the concession of a far earlier development of that division of national civilization, which is comprised in the "art of writing,” than the majority of Palæographers are prepared to recognize. Let Hieroglyphics and Cuneiform retain their ancient fame; but the question succeeds, as to how close upon their earliest traces did other systems of writing assert themselves, more facile in materials and more suitable for the purposes of commercial and private life than the formal sculptured figures of the Egyptian temples, or the complicated arrowheaded syllabary of Mesopotamian Palaces, which latter mechanism, however, in its transitional variations, so firmly retained popular favour in virtue of its applicability to the ever-ready clay, the comparative indestructibility of which had been established by many ages of local use.1

Egyptologers, on their part, concede a very archaic date for the use of parallel systems of writing, and the age of Phoenician, with our present information, need no longer be narrowed within the limits defined by its surviving monuments, the majority of which must be held to have disappeared with the perishable material chiefly used for their reception. It it is clear that some form of Phoenician, constituting a kind of current hand, was in official use under the Assyrian kings, as the authoritative definition of the lion-weights in the letters of that alphabet sufficiently declares; and we are further justified in assuming, in all cases where two Scribes are represented in the royal sculptures, that in intentional contrast to the Cuneiform manipulator, the second amanuensis, who uses a reed and a parchment

1 Rawlinson, J.R.A.S. x, pp. 32, 340, and vol. i. N.S. p. 245. See also the names of Seleucus Philopater (187-175 B.C.), Antiochus (175-164 B.C.), and Demetrius (146-139 B.c.), upon the Cuneiform tablets of terra-cotta in the British Museum, deciphered by Oppert, "Expédition en Mésopotamie,” ii. 357.

roll, is designed to portray a man writing with ink in some one of the, as yet, but slightly divergent provincialisms of archaic Phoenician.

Sargon's Record Chamber has already proved itself a perfect storehouse of palæographic data, and, if I am not mistaken, may claim to add another to its list of contemporary alphabets. Mr. Layard, in his admirable description of his own discoveries at Koyunjuk, interested his readers in an unusual degree by an account of the still surviving association of the hieroglyphic signet of Subaco, with that of the Assyrian king on a lump of clay, which was supposed to have formed the connecting attestation of the less permanent substance upon which some royal treaty or compact had been engrossed. In the same closet were found several impressions of smaller seals on suitably-sized bits of clay, which at the time attracted no attention; these, however, on closer scrutiny, seem to bear four varying letters, which can scarcely represent anything but ancient Ethiopian characters; at least two, if not three out of the four letters are readily identifiable with certain corresponding characters of the modern alphabets. It is not necessary, for the purpose of proving the currency of this form of writing, that we should be able to detect any of the leading names, either of Subaco, his relatives, or ministers. The importance of the identification consists in the very unexpected determination of the definite antiquity of the writing of the Ethiopian and cognate nationalities, and the very close bearing this date has upon the alphabetical schemes of the

1 Mr. Layard's account of the discovery of these seals is as follows:-"In a chamber or passage [leading into the archive chamber] in the south-west corner of the palace of Kouyunjik, were found a large number of pieces of fine clay bearing the impressions of seals, which, there is no doubt, had been affixed, like modern official seals of wax, to documents written on leather, papyrus, or parchment. Such documents, with seals of clay still attached, have been discovered in Egypt, and specimens are still preserved in the British Museum. The writings themselves have been consumed by the fire which destroyed the building or had perished from decay. In the stamped clay, however, may still be seen the holes for the string or strips of skin by which the seal was fastened; in some instances the ashes of the string itself remain, with the marks of the fingers and thumb. The greater part of these seals are Assyrian; but with them are others bearing Egyptian, Phoenician, and doubtful symbols and characters. But the most remarkable and important of the Egyptain seals are two impressions of a royal signet, which, though imperfect, retain the cartouche, with the name of the king, so as to be perfectly legible. It is one well known to Egyptian scholars as that of the second Sabaco, the Ethiopian of the twenty-fifth dynasty. On the same

Indian Ethiopians, and the kindred nations to the southeastward, in which many points of constructive identity have already been recognized.

piece of clay is impressed an Assyrian seal, with a device representing a priest ministering before the king, probably a royal signet."

The annexed woodcut outlines represent six of the Ethiopian seals, copied from the extant clay-impressions of the original signets, that have survived both "Nineveh and Babylon." My object in this, and I trust in all similar cases, is not to force

identities, but to place before my fellow labourers coincidences that may perchance elicit new truths. It is not pretended that the literal symbols here found associated with Egyptian hieroglyphics and Assyrian cuneiform will tally or accord exactly with the transmutations incident to the alphabetical developments of the once powerful, but for many centuries obscure, nationalities that in the interval must have remained more than ordinarily indebted to the advancing world around them. Under this latitude of identification, we may freely appeal to the later forms of Ethiopic, Amharic, or other cognate conservators of traces of the ancient writing, though it is more to the general palæographic configuration than to absolute and complete uniformity of outline that any test must be applied.

It may be said in regard to the seals now presented, that they convey in all but five independent letters; the most marked of the number is the , which occurs

with sufficient clearness on three occasions. There can be little hesitation in asso

ciating this form with the modern Himyaritic

especially when the subjunct vowel i is added,

sh or the Ethiopian † shă,

which is so distinctly seen in

a varied form, even under possible repetition, in the ancient example. The second figure of special mark is the, which offers a more dubious range of identification among the derivative Ethiopian forms of bi, ʼn bě, extending even to the Amharic ʼn khă, and many other possible renderings; but the most curious coincidence is in the near connection of the sign with the Sanskrit of Northern India (Prinsep's Essays, ii. p. 40, pl. xxxviii.).

The third character, which almost seems to have been in a transition stage at the time these seals were fashioned, may be reduced in the modern alphabets to the Ethiopian ( tă or 00 mă; but of the prevailing coincidencies of formation under the general Ethiopian scheme there can be little question.

The imperfect outline , which recurs on four occasions, may be an Amharic Hjă, or other consonantal combination of j, with a different vowel: an approximate likeness is also to be detected to the Coptic X j; or the old figure may, perchance, constitute the prototype of the modern Himyaritic

m.

1 Herodotus, ii. 94; vii. 70. Rawlinson's Herodotus, vol. i. 650; iii. 264, note 1; iv. p. 220. J. R. A. S. xv. 233.