Position sense of the peripheral retina.

Position acuity was measured over a wide range of eccentricities, from 3 min to 10 deg in the horizontal meridian, by using both a three-dot bisection task and a three-dot vernier task. A foveal fixation dot appeared for 1 sec before an outer pair of test dots was flashed for 200 msec. Bisection and vernier tasks were used to measure position acuity in the radial and tangential directions, respectively. The vernier data were well fitted by a straight line on linear axes of offset threshold versus eccentricity. The bisection data, on the other hand, were poorly fitted by a single straight line. However, a double-line fit worked very well. The line segment at large eccentricities (greater than 0.5 deg) had an x intercept of about 0.6 deg, in good agreement with previous estimates based on cortical magnification and on hyperacuity in the presence of flanks. These results imply that three-dot vernier thresholds are set by a single orientation mechanism at all eccentricities and that three-dot bisection thresholds are set by a pair of mechanisms. For eccentricities less than 15 min, thresholds are in good agreement with calculations based on spatial-frequency filters. For larger eccentricities, the bisection thresholds agree with scaled anatomical modules that are presumed to exist in the human visual cortex. The thresholds for position acuity in the tangential direction are as low as 0.005 times the eccentricity. In the radial direction, thresholds are poorer, implying that additional cortical factors may further constrain performance.