Rawlinson had worked at it for some time, he found that some of these were only variants, or slightly deviating forms of the same letter; but having discovered this, and determined the value of the alphabetic letters, the language still remained to be mastered. An unexpected aid was about this time discovered. Just as Arab, Persian, and Turk, exist side by side in Mesopotamia at the present day, so did the Assyrian, the Persian or Mede, and the Scythian, in the days of Darius. To this circumstance we owe it that any progress has been made in their decipherment. All of them are trilingual: one written in Persian, another in Assyrian, and a third in a language which has not yet been fully deciphered. The Behistun inscription from which Colonel Rawlinson picked out his Assyrian contains from 80 to 100 proper names, which he could now read in the Persian cuneiform writing; it was, therefore, not difficult to construct an Assyrian alphabet pretty nearly accurate. The most frequently recurring words were soon recognised; and when the sound had been approximatively determined, it was found that the language was very nearly allied to the Hebrew and the ancient Chaldee. It will not be supposed that, even after this discovery, Colonel Rawlinson's task was henceforth easy. Obstacles lay in his way, of which students who learn a language with all the aids of lexicons, grammars, and annotated texts, have no conception. Thus, this Behistun inscription is engraved on a rock at an elevation of 300 feet above the plain; and its delicatelyexecuted characters had to be read by the aid of a telescope; besides which, a part of it was peeled off and irrecoverably lost. The inscriptions at Persepolis were so short, so crowded with proper names, and so full of repetition, that it was difficult to ascertain what the real language was. In spite of all these impediments, Colonel Rawlinson considers the meaning of about 500 words as certainly determined; and as these contain many substantives, verbs, and adjectives, with probably all the prepositions, they suffice to explain the meaning of any simple record of events, and such is the character of most of these inscriptions.

The inscriptions at Khorsabad are never found upon any of the façades, but run along the sides of the chambers, forming a line between the upper and lower bas-reliefs. There are also shorter ones engraved upon the bottom of the dresses of

470

INSCRIPTIONS.-RAWLINSON.

the different figures, and others still briefer between the legs of the bulls at the doorways, as well as on the large flags which pave the entrance to the doors. Besides these, others, seemingly consisting of a single word, are to be seen over the heads of captives, and the representation of different towns. These Botta conjectures to be proper names. Another class of inscriptions was discovered upon the back of the gypsum slabs which formed the panelling of the chambers. Botta at first accounted for this fact by supposing that the remains of some still more ancient building had been employed in the construction of the Khorsabad monument; but as the inscriptions were always the same, and invariably placed in the very middle of the block, he came to the conclusion that they must represent the name or genealogy of the monarch who raised the structure, or else commemorate some historical fact. This supposition is strengthened by the circumstance that the inscriptions in question are also cut upon the sides of the stones which formed the angle of the chambers. They were not executed with the same care and nicety as those upon the walls of the chambers, but were evidently placed in the position they occupied, in the same manner, and for the same reasons, that coins and medals are deposited under the foundation-stones of modern buildings.

The inscriptions at Khorsabad are, without exception, all written in the cuneiform character, and, with few variations, the same as that employed at Nimroud. This fact fixes the date of the monument anterior to the termination of the Assyrian empire. Botta gives, at great length, a catalogue of the characters he met with at Khorsabad, and also a list of the different groups formed by these simple characters or elements, and finds these groups, including the variations which he observed in their form, to amount to 642. The number of simple elements in each group varied from one to fourteen, but never exceeded the latter number. Botta is of opinion that the different groups are not resolvable into their simple elements, but that each represents a separate sound, as in Chinese: in this view he differs from all other inquirers. At Khorsabad a great many inscriptions illustrate historical subjects, and it cannot be supposed that they always contain the same individual words. With so small a number of groups, therefore, it is impossible each group can have represented a word; they

must evidently stand for either a letter or a syllable. The words, too, generally consist of a number of signs or groups, varying from one to four, from which it may be concluded that the language is syllabic, or that, at least, the signs representing the consonants contain also the necessary accompaniment of vowels. Botta was at first inclined to believe in the co-existence of another system of writing, on account of the complexity of the cuneiform, and also because he discovered bricks, vases, and gems, with inscriptions somewhat resembling the Phoeni cian character. He accounts for this, however, by supposing that the cuneiform letters may, like the Chinese, for ordinary use, be written quickly, and, as is the case with hieroglyphics, be reduced to such simplicity as to become almost irrecognisable as variants of the normal form. He also suggests, as a reason for the two systems of writing, that as the Phoenicianlike characters were always found upon small articles, such as gems, vases, cylinders, &c., they might have been the work of foreign workmen, anxious to leave some mark of their nationality, or may have been engraved by the captives who were kept prisoners by the monarchs of Assyria. This may certainly have been the case at Babylon, where many of these objects with the inscriptions in question were discovered, and where there was a constant communication with the Phoenician populations inhabiting the shores of the Mediterranean.

There is one remarkable fact connected with the cuneiform inscriptions of Khorsabad. No modification ever, or hardly ever, is observable at the commencement, or in the middle of the words. The termination alone is affected. This peculiarity, Botta thought, went far to prove that the language was not Semitic, as in the latter class of languages the changes always occur in the beginning; nor is it of the Arian family, as there are no traces of prefixed prepositions or composed words.

Having given, we trust, full credit to the acumen of Grotefend and to the profound learning and skill of Lassen, we may now devote the remainder of our space to an account of the labours of our own countryman, Rawlinson, of whom every Englishman may well be proud. We shall do this chiefly in his own words, as contained in the "Journal of the Asiatic Society." In a memoir, prepared in 1839, but not then published, the

472

INSCRIPTIONS.-RAWLINSON.

Colonel thus wrote respecting the labours of his predeces

sors:

"It would be interesting, perhaps, to the lovers of Oriental literature, if I could open the present memoir with a detailed account of the progress of cuneiform discovery, from the time when Professor Grotefend first deciphered the names of Cyrus, Xerxes, and Darius, to the highly improved condition which the inquiry now exhibits; but my long absence from Europe, where the researches of Orientalists have been thus gradually perfecting the system of interpretation, while it has prevented me from applying my own labours to the current improvements of the day, has also rendered me quite incompetent to discriminate the dates and forms under which these improvements have been given to the world. The table, however, in which I have arranged the different alphabetical systems adopted both by continental students and by myself, will give a general view of their relative conditions of accuracy, and supposing the correctness of my own alphabet to be verified by the test of my translations-it will also show that the progress of discovery has kept pace pretty uniformly with the progress of inquiry.

"Professor Grotefend has certainly the credit of being the first who opened a gallery in this treasure-house of antiquity. In deciphering the names of Cyrus, Darius, Xerxes, and Hystaspes, he obtained the true determination of nearly a third of the entire alphabet, and thus at once supplied a sure and ample basis for further research. M. Saint Martin, who resumed the inquiry on its being abandoned by the German professor, improved but little on the labours of his predecessor: but shortly afterwards Professor Rask discovered the two characters representing M and N, which led to several most important verifications.

"The memoir of M. Bournouf on the two cuneiform inscriptions of Hamadán, published in 1836, added several discoveries of interest; and the recent researches of Professor Lassen, supplying an identification of at least twelve characters, which had been mistaken by all his predecessors, may entitle him almost to contest with Professor Grotefend the palm of alphabetical discovery..

"In a very few cases only, which may be seen on a reference to the comparative table, have I indeed found occasion to differ

with him as to the phonetic power of the characters, and in some of the cases even, owing to the limited field of inquiry, I have little more than conjecture to guide me.

"But in thus tracing the outlines of the discovery as far as they are at present known to me, and in thus disclaiming any pretension to originality as far as regards the alphabet which I have finally decided on adopting, I think it due to myself to state briefly and distinctly how far I am indebted for my knowledge of the cuneiform character, and of the language of the inscriptions, to the labours of continental students which have preceded the present publication. It was in the year 1835 that I first undertook the investigation of the cuneiform character. I was at that time only aware that Professor Grotefend had deciphered some of the names of the early sovereigns of the house of Achæmenes; but in my isolated position at Kermanshah, on the western frontier of Persia, I could neither obtain a copy of his alphabet, nor could I discover what particular inscriptions he had examined. The first materials which I submitted to analysis were the sculptured tablets of Hamadán, carefully and accurately copied by myself upon the spot; and I afterwards found that I had thus, by a singular accident, selected the most favourable inscriptions of the class which existed in all Persia for resolving the difficulties of an unknown character.

"These tablets consist of two trilingual inscriptions, engraved by Darius Hystaspes and his son Xerxes. They commence with the same invocation to Ormazd (with the exception of a single epithet omitted in the tablet of Darius); they contain the same enumeration of the royal titles, and the same statement of paternity and family; and, in fact, they are identical, except in the names of the kings and in those of their respective fathers. When I proceeded, therefore, to compare and interline the two inscriptions (or rather, the Persian columns of the two inscriptions; for as the compartments exhibiting the inscription in the Persian language occupied the principal place in the tablets, and were engraved in the least complicated of the three classes of cuneiform writing, they were naturally first submitted to examination), I found that the characters coincided throughout, except in certain particular groups, and it was only reasonable to suppose that the groups which were thus brought out and individualised must