"It was clear that the three versions of the Achaemenian inscriptions were addressed to the three chief populations of the Persian empire, and that the one that invariably came first was composed in ancient Persian, the language of the sovereign himself. Now this Persian version happened to offer the decipherer less difficulties than the two others which accompanied it. The number of distinct characters employed in writing it did not exceed forty, while the words were divided from one another by a slanting wedge. Some of the words contained so many characters that it was plain that these latter must denote letters, and not syllables, and that consequently the Persian cuneiform system must have consisted of an alphabet, and not of a syllabary. It was further plain that the inscriptions had to be read from left to right, since the ends of all the lines were exactly underneath one another on the left side, whereas they terminated irregularly on the right; indeed, the last line sometimes ended at a considerable distance from the right-hand extremity of the inscription. "The clue to the decipherment of the inscriptions was first discovered by the successful guess of a German scholar, Grotefend. Grotefend noticed that the inscriptions generally began with three or four words, one of which varied, while the others remained unchanged. The variable word had three forms, though the same form always appeared on the same monument. Grotefend, therefore, conjectured that this word represented the name of a king, the words which followed it being the royal titles. One of the supposed names appeared much oftener than the others, and, as it was too short for Artaxerxes and too long for Cyrus, it was evident that it must stand either for Darius or for Xerxes. A study of the classical authors showed Grotefend that certain of the monuments on which it was found had been constructed by Darius, and he accordingly gave to the characters composing it the values required for spelling 'Darius' in its old Persian form. In this way he succeeded in obtaining conjectural values for six cuneiform letters. He now turned to the second royal name, which also appeared on several monuments, and was of much the same length as that of Darius. This could only be Xerxes; but if so, the fifth letter composing it (r) would necessarily be the same as the third letter in the name of Darius. This proved to be the case, and thus afforded the best possible evidence that the German scholar was on the right track. "The third name, which was much longer than the other two, differed from the second chiefly at the beginning, the latter part of it resembling the name of Xerxes. Clearly, therefore, it could be nothing else than Artaxerxes, and that it actually was so was rendered certain by the fact that the second character composing it was that which had the value of r.