In response to Cohn's 1989 remarks, we believe his view that cautious interpretation of Grosser and Spafford's 1989 findings is based on an overly hasty reading of our work. We address his issues point by point. Our original study hypothesizes that an unusually high cone density occurs in the periphery of the dyslexic's eye; such a hi h density would account for the high visual acuity noted by Geiger and Lettvin (1987). Even if normal readers' eyes have peripheral cones, these are apparently not arranged densely enough to provide such keen, acute detail vison. Problem I. The pothesis.-The present authors do not declare the peripheral retinas of 2 human eyes to be tot y void of cones. It is a well-known fact that, although the human retina contains approximately 6 million cones (Kalat, 1984, p. 156), these are more commonly located toward the center of the retina (Kalat, 1984, p. 129). Cohn attributes fixation deviations to variations in fixation by the dyslexic reader, particularly along the horizontal meridian. Yes, fixation deviations are relatively large along the horizontal meridian. However, Curcio, Sloan, Packer, Hendrickson, and Kalina (1987) point out that, in their human observers, cones tend to drop more slowly in density along the horizontal meridian than they do along the vertical meridian. If dyslexic eyes tend to lose cone density more slowly in peripheral directions, relatively greater numbers of extra cones should occur along the horizontal than the vertical meridian. Specifically, the sharp decline in cone density from 0 to lo0 indicated in the Osterberg figure (cited by Cohn, 1989) may not (in our view) occur in the dyslexic retina. Perhaps in dyslexics' retinas, the iso-density line runs horizontally across the top of the graph (50,00O/sq. mm.) or perha s the density function decreases slowly from O0 to lo0. rolle em II: Pivial alternative Aypotheses.-These comments do not require response. Problem HI: Reqzriremenk for psychophysical studies of photoreceptors are not met.-If Cohn is correct in assuming the ability of d slexics to identify colored targets is due to reflected energies using their rods, then why was tLs ability not evidenced in the case of proficient readers? Should not their rods [unction also? Concluding remarks.-Cohn, based on his various objections, suggests work along this Line is not warranted. The attempt to shut down further work along the lines indicated by our research seems counter-productive. At the moment, clinical activities aimed at relieving reading difficulties have very little theoretical backing. The building of a sound theoretical understanding can only be helpful. Setting priorities for topics seems preferable to cessation of research efforts. The existence of methodological problems is what makes science interesting and challenging. We must be ingenious to make discoveries in the face of technical difficulties.