sharpers and political mountebanks he will look up to and try to imitate all the rest of his life.
So far as I can discover, there has been but one attempt by a competent authority to determine the special characters of this general tongue of the mobile vulgus. That authority is Dr. W. W. Charters, now head of the School of Education at the University of Illinois. In 1914 Dr. Charters was dean of the faculty of education and professor of the theory of teaching in the University of Missouri, and one of the problems he was engaged upon was that of the teaching of grammar. In the course of this study he encountered the theory that such instruction should be confined to the rules habitually violated—that the one aim of teaching grammar was to correct the speech of the pupils, and that it was useless to harass them with principles which they already instinctively observed. Apparently inclining to this somewhat dubious notion, Dr. Charters applied to the School Board of Kansas City for permission to undertake an examination of the language actually used by the children in the elementary schools of that city, and this permission was granted. The materials thereupon gathered were of two classes. First, the teachers of grades III to VII inclusive in all the Kansas City public-schools were instructed to turn over to Dr. Charters all the written work of their pupils, "ordinarily done in the regular order of school work" during a period of four weeks. Secondly, the teachers of grades II to VII inclusive were instructed to make note of "all oral errors in grammar made in the school-room and around the school-building" during the five school-days of one week, by children of any age, and to dis- patch these notes to Dr. Charters also. The result was an accumulation of material so huge that it was unworkable with the means at hand, and so the investigator and his assistants reduced it. Of the oral reports, two studies were made, the first of those from grades III and VII and the second of those from grades VI and VII. Of the written reports, only those from grades VI and VII of twelve typical schools were examined.
The ages thus covered ran from nine or ten to fourteen or fifteen, and perhaps five-sixths of the material studied came from
