Twelve years of his life are said to have been given to his great work. Being entirely without instruction and having no knowledge of the philosophy of language, being not even acquainted with English, his first attempts were naturally enough in the direction of the crude Indian pictograph. He set out to devise a symbol for each word of the language, and after several years of experiment, finding this an utterly hopeless task, he threw aside the thousands of characters which he had carved or scratched upon pieces of bark, and started in anew to study the construction of the language itself. By attentive observation for another long period he finally discovered that the sounds in the words used by the Cherokee in their daily conversation and their public speeches could be analyzed and classified, and that the thousands of possible words were all formed from varying combinations of hardly more than a hundred distinct syllables. Having thoroughly tested his discovery until satisfied of its correctness, he next proceeded to formulate a symbol for each syllable. For this purpose he made use of a number of characters which he found in an old English spelling book, picking out capitals, lower-case, italics, and figures, and placing them right side up or upside down, without any idea of their sound or significance as used in English (see plate v). Having thus utilized some thirty-five ready-made characters, to which must be added a dozen or more produced by modification of the same originals, he designed from his own imagination as many more as were necessary to his purpose, making eighty-five in all. The complete syllabary, as first elaborated, would have required some one hundred and fifteen characters, but after much hard study over the hissing sound in its various combinations, he hit upon the expedient of representing the sound by means of a distinct character--the exact equivalent of our letter s--whenever it formed the initial of a syllable. Says Gallatin, "It wanted but one step more, and to have also given a distinct character to each consonant, to reduce the whole number to sixteen, and to have had an alphabet similar to ours. In practice, however, and as applied to his own language, the superiority of Guess's alphabet is manifest, and has been fully proved by experience. You must indeed learn and remember eighty-five characters instead of twenty-five [sic]. But this once accomplished, the education of the pupil is completed; he can read and he is perfect in his orthography without making it the subject of a distinct study. The boy learns in a few weeks that which occupies two years of the time of ours." Says Phillips: "In my own observation Indian children will take one or two, at times several, years to master the English printed and written language, but in a few days can read and write in Cherokee. They do the latter, in fact, as soon as they learn to shape letters. As soon as they