The direct examination of the individual should be supplemented by a family history, a development history, and a school history. In some cases special deficiency in spelling seems to be hereditary. Earle has made a study of the inheritance of capacity for spelling, from which he concludes that there is distinct fraternal resemblance in spelling. Stephenson has reported six cases of special inability to read and spell, which occurred in three generations of one family. IV. CAN SPECIAL DEFECT IN SPELLING BE OVERCOME? Spelling has received relatively little study as a process, in comparison with the attention which has been given to reading and arithmetic. We have no variety of experiments carried out to improve poor spellers, as we have in the case of poor readers. In 1918 the present writer reported, with Miss Winford, the results of studying and teaching a group of poor spellers, from the fifth grade. The experiment extended over two periods of ten weeks each, but the time was largely devoted to observations of the errors made, measurements of intelligence, and inventions of incentives for arousing interest in spelling as a group project. No child was taken individually, and given intensive instruction, as with the boy, X, in reading, reported in Chapter IV. During the period of class teaching, all the poor spellers improved, as measured by the Ayres scale, but the three very poorest still remained at the bottom of the class. By intensive individual instruction any one of these three might have made much greater improvement. We are, therefore, now in need of experiments carried out to improve poor spellers. Such experiments must include precise measurements of intelligence, ability to spell, ability to read, and amount of time expended. They must include a description of the sensory equipment of the spellers, and information on all points listed under the suggested outline for the examination of poor spellers. There must be an adequate account of method used, and objective measurements of improvement must be presented. From knowledge of spelling as a process of habit formation, it would be predicted that any child of average intelligence, and normal sensory capacity, can learn to spell, if sufficient drill be undergone. English spelling is, however, relatively resistant to learning, because of the specific character of the connections to be made. Very few generalizations are possible, each word being to so great an extent a special matter. For this reason it is very important to teach first the words most commonly used. These have been ascertained by research in the Russell Sage Foundation. V. DOES READING TEACH SPELLING?