“…However, a variation of 0n27 mm in the radius of curvature represents a change of 0n25 D in refraction when the cornea is in contact with the aqueous humor. This level of error would be within the limit of reproducibility of modern videokeratoscope (McCarey, Zurawski and O'Shea, 1992 ;Douthwaite, 1995 ;Douthwaite and Matilla, 1996). Considering the satisfactory repeatability of the measurements shown by the current tests on calibration lens and normal human eyes, we suppose that the relative values like astigmatism obtained herein would hold reasonable accuracy even if some doubt may exist about the absolute measurement values of the videokeratoscope (Douthwaite, 1995 ;Douthwaite and Matilla, 1996).…”