The children, however, must have been well taught. This is clear from the remarkably good spelling which we find in the private letters; it is seldom that words are misspelt. The language may be conversational, or even dialectic, but the words are written correctly. The school-books that have survived bear testimony to the attention that had been given to improving the educational system. Every means was adopted for lessening the labor of the student and imprinting the lesson upon his mind. The cuneiform characters had been classified and named; they had also been arranged according to the number and position of the separate wedges of which they consisted. Dictionaries had been compiled of Sumerian words and expressions, as well as lists of Semitic synonyms. Even grammars had been drawn up, in which the grammatical forms of the old language of Sumer were interpreted in Semitic Babylonian. There were reading-books filled with extracts from the standard literature of the country. Most of this was in Sumerian; but the Sumerian text was provided with a Semitic translation, sometimes interlinear, sometimes in a parallel column. Commentaries, moreover, had been written upon the works of ancient authors, in which difficult or obsolete terms were explained. The pupils were trained to write exercises, either from a copy placed before them or from memory. These exercises served a double purpose—they taught the pupil how to write and spell, as well as the subject which the exercise illustrated. A list of the kings of the dynasty to which Khammurabi belonged has come to us, for instance, in one of them. In this way history and geography were impressed upon the student’s memory, together with extracts from the poets and prose-writers of the past. The writing material was clay. Papyrus, it is true, was occasionally used, but it was expensive, while clay literally lay under the feet of everyone. While the clay was still soft, the cuneiform or “wedge-shaped” characters were engraved upon it by means of a stylus. They had originally been pictorial, but when the use of clay was adopted the pictures necessarily degenerated into groups of wedge-like lines, every curve becoming an angle formed by the junction of two lines. As time went on, the characters were more and more simplified, the number of wedges of which they consisted being reduced and only so many left as served to distinguish one sign from another. The simplification reached its extreme point in the official script of Assyria.