Reading ability is no longer considered as one general ability.' The general ability has been minutely analyzed2 into a large number of very specific and distinct skills, each one of which may function in definite reading situations. Attempts have been made to designate the school grades in which certain of these skills should be taught.3 Standardized reading tests have been constructed largely on the assumption that the ability to read consists in but one general ability. If such a test is given to a group of pupils, there is no certain way of knowing which of the many specific reading skills have been tested. The resulting scores may indicate whether a pupil is above, at, or below the norm for his age and grade in the reading skills examined by that test. The scores do not indicate the pupil's relative rank in other reading skills. If a pupil scores decidedly below the norm for his age and grade on several of the available standardized reading tests, what does that mean? What can and should be done to improve that pupil's reading skills? The following is a description of an experiment in which six over-age fourth-grade pupils of low achievement in reading were given intensive instruction in work-type reading. Table I shows the chronological ages, mental ages, intelligence quotients, and reading scores of these six pupils. The pupils were from thirteen to twenty-seven months over-age chronologically. Three of them were also over-age mentally. According to their men-