In the literature on the subject little reference is made to the relation between transfer of training and intelligence, and no study of it seems to have made use of the intelligence-testing technique developed in the last few years. 1 Judd (1908), primarily interested in the role of generalized experience, discussed the Relation of Special Training to General Intelligence, but no intelligence tests were used, the first revision of the so-called 1905 Binet Scale having just been issued. Rugg (1916) analyzed and summarized the studies of transfer up to that time. In his own study, he found transfer of training in descriptive geometry positively related to scholarship in "disciplinary or training courses, such as the various college courses in mathematics," but neutrally related to scholarship in English and modern languages.To determine the amount of transfer from training in mental multiplication and substitution and its relation to intelligence, the following investigation was made in an elementary school in Baltimore in May and June, 1923. One hundred sixty pupils of Grades Via, Vila, VIII6, and Villa were selected to form the experimental and control groups.Test Series I (T.S.I.) in Grades Vila to Villa consisted of the Courtis Arithmetic, Form B (division being given as mental division with a 10-minute time allowance), and the following tests devised by the writer: Two French vocabulary tests, two letter-digit substitution, one number cancellation, one a-t cancellation, a modified form of Binet's simultaneous adding, immediate memory for words and digits (three-to eight-term, and three-to ten-term lists), drawing lines equal to 21 given lines of various lengths from % to 4 in., arranged in mixed order, and bisecting the given lines. Test Series I for the Via classes included all of the tests for Grades VII and VIII except the substitution and a-t cancellation tests. Courtis Division was not given as mental division, but according to Courtis' directions. 1 In the February number of this Journal (pp. 90, 97) Thorndike uses .09 as a conservative estimate of the true correlation between initial ability (in intelligence tests) and gains for pupils taking the same courses. Gain probably includes (a) a year's general mental growth, (b) practice effect in taking the tests, and (c) special (or general) training from the school studies. 4K5