Reading disability is one of the foremost problems of elementary edu cation. It consists of the inability to learn to read normally, even when the pupil is of average or superior intelligence. This disability has often been mistaken for dullness and the pupil has been side-tracked into some occupation not requiring much intelligence. This frequently has resulted in a loss of service to the community, while the individual has been de prived of certain intellectual pleasures and associations which might have meant much to him. Too little is known of the etiology and treat ment of reading difficulties, and definite information of this kind is urgently needed. Reading is presented through visual channels, and a very natural in ference is that if the visual system is defective reading must be impeded. Accordingly a study of the ocular characteristics of two groups of chil dren has been made. One group was composed of one hundred fourteen reading disability cases, and the other consisted of one hundred forty three unselected school children. Included in these totals are fifty-six unselected cases which were obtained through the cooperation of Super intendent Frank A. Scott and Miss Eleanor Hayes of the Belmont, Massachusetts, School Department. Most of the one hundred fourteen reading disability cases were taken from my private practice and I exam ined the others at the Harvard Psycho-Educational Clinic. Sixty-four of the reading disability, and eighty-seven of the unselected cases were studied separately and were reported in the Eye, Ear, Nose and Throat Monthly of April, 1931. Visual acuity was measured monocularly with a standard Snellen chart at twenty feet. Coordination was investigated by means of a num ber of ophthalmic instruments, among which were the Stevens phorometer, Risley rotary prisms, Maddox rods, and double prisms. Adduction and abduction were measured by means of the rotary prisms. The use of a Wells-DeZeng phoroptor increased the accuracy of the measurements. In a previous study the median refractive errors exhibited by both the reading disability and unselected groups differed so slightly that no actual measurements of the degree of refractive defect were employed in this 211