date had noticed these inscriptions at Persepolis and elsewhere, and while some compared the forms of the characters composing them to arrows, others considered them to be wedges, cunei in Latin. The latter comparison was the origin of the term 'cuneiform,' ordinarily applied to them. We find it already used by Hyde in his Historia Religionis veterum Persarum, which was published at Oxford in 1700[2].

The Italian traveller, Pietro della Valle, in 1621, was the first who made the characters known in Europe by printing a few of them; at the same time he put forward the correct suggestion that the inscriptions were to be read from left to right. A more important collection of signs, however, was published in 1693, in one of the early volumes (No. 201) of the Philosophical Transactions of the Royal Society from the papers of Mr. Flower, who had been specially charged by the East India Company with the duty of investigating the antiquities of Persia. But it was not till the middle of the eighteenth century that Cornelius van Bruyn (1714) and Carsten Niebuhr (1774-8), the father of the historian, first copied and published the inscriptions in anything like a complete and accurate manner. Niebuhr further pointed out that they comprised three different systems of cuneiform writing, which in the case of every text followed one another in a regular order. The first system of writing was the simplest, as it consisted of only forty-two different characters, whereas the number of characters in the second and third systems was very large.

With Niebuhr's publication the work of decipherment became possible. In 1798, Professor Tychsen, of Rostock, discovered that in the first system an oblique wedge was used to divide the words from one another, and in 1802 the Danish Bishop, Münter, starting from this basis, showed that the language possessed suffixes, pointed out that certain characters denoted vowels, and even divined the word for 'king,' as well as the value of two letters, one of them being a. He also maintained that while the first system of writing was alphabetic, the second was syllabic, and the third ideographic, and that as the inscriptions were found in Persia and on the buildings of the Achaemenian kings, the text which always comes first must represent the language of ancient Persia, which he identified, though erroneously, with Zend.

It is, however, to George Frederick Grotefend, of Hanover, that the discovery of the key which has unlocked the secrets of cuneiform literature is really due. On September 4, 1802, he read before the Royal Society of Göttingen a Memoir, in which he announced his discovery of the names of certain Achaemenian kings in the cuneiform inscriptions, and explained the method by which he had arrived at his results. By a curious coincidence it was at the same meeting of the Society that Heyne described the first efforts that had been made towards deciphering the Egyptian hieroglyphics. Grotefend first showed convincingly that the inscriptions must be read from left to right, a portion of a word which ends a line on the right side in one of the texts beginning the next line on the left side in a duplicate copy of it. He next pointed out that the analogy of the Sassanian inscriptions, which had just been deciphered by de Sacy, indicated that the Persepolitan texts must commence with the names of the kings who had erected the monuments, followed by their titles, and that a comparison of the texts one with another made it pretty evident that such was actually the case. In this way he succeeded in finding (like Münter before him) the word for 'king,' and in addition to this the royal names preceding it. Those on the Persepolitan monuments represented a father and a son, though in certain cases the father added his own father's name, but without the royal titles. Thanks to the classical writers, it was known that the monuments were of Achaemenian origin, and the names of the Achaemenian kings had also been preserved. It only remained to fit them to the characters in the cuneiform texts. Hystaspes, Darius, and Xerxes alone suited, since Cyrus was too short and Artaxerxes too long; moreover, the letters a, r, and sh, in the names of Darius and Xerxes appeared in their right places if these names were adopted. So, too, did a and sh in the name of Hystaspes. Such a coincidence was sufficient to prove that Grotefend was right in his guess that the words in question represented proper names, for guess it was, though founded on strong probability and scientific induction. He had noticed that two of the names (those of Darius and Xerxes) occurred separately on two particular groups of monuments, whereas the word which followed them was always the same. It was natural to conclude that the latter word denoted 'king,' while those which preceded it were proper names.

The alphabet Grotefend had constructed out of the proper names enabled him to read the word for 'king,' and thus to show its near affinity to the corresponding word in Zend. But he was a classical scholar rather than an orientalist, better known by his Latin grammar than by his knowledge of Eastern languages, and consequently as soon as his pioneering work of decipherment was accomplished, he lacked the philological knowledge which would have allowed him to continue it. Moreover, he was hampered by the false theory that the language of the inscriptions was identical with Zend. The next step of importance was taken by Rask in 1826, who discovered the termination of the genitive plural and the true reading of the title 'Achaemenian.' Rask was followed in 1836 by the great Zendic scholar Burnouf at Paris, and by Lassen at Bonn. Burnouf demonstrated that the language of the Achaemenian texts was not Zend, but a sister dialect spoken in western Persia, and his discovery of the names of the satrapies, in one of the inscriptions copied by Niebuhr, enabled him and Lassen simultaneously almost to complete what we may henceforth call the Old Persian alphabet. A few corrections in it were subsequently made by Beer, Jacquet, Holzmann, and Lassen himself.

Meanwhile a young English officer in the East India Company's service, now Sir Henry Rawlinson, had been working in Persia unassisted, and at a distance from libraries, upon the Old Persian texts. He knew that Grotefend had discovered in them the names of the early Achaemenian monarchs, and with this clue he set himself to construct an alphabet and interpret the inscriptions. He soon found means of providing himself with fuller materials for the work of decipherment than those at the disposal of scholars in Europe, by copying the great inscription which Darius had caused to be engraved on the sacred rock of Bagistana or Behistun in commemoration of his accession to the throne of Persia, and re-conquest of the empire of Cyrus. The task of copying the inscription—by far the longest Persian one known—was an arduous one, and not unattended with danger, and it occupied several years. Rawlinson first saw the inscription in 1835; it was not till 1839 that the whole of it was copied. A few years later he revised it again, but his memoir upon it and upon the other Old Persian texts was not ready for publication till 1845. In the following year the text was published by the Royal Asiatic Society, and the translation and commentary followed in 1849. Dr. Hincks, of Dublin, had already (in 1846) given the last touch to the decipherment of the Old Persian alphabet by the discovery that the consonants composing it contained inherent vowels.

As we have seen, Niebuhr had perceived that the Persepolitan inscriptions were in three different systems of writing. But it was only after the decipherment of the Persian texts that it was found that the three systems of writing embodied three separate languages, and belonged to three separate countries. As in modern Turkey a