J. Physiol. (1974), 243, pp. 739-756 739 With 9 text-figures Printed in Great Britain THE EFFECT OF EARLY ASTIGMATISM ON THE VISUAL RESOLUTION OF GRATINGS By DONALD E. MITCHELL AND FRANCES WILKINSON From the Department of Psychology, Dalhousie University, Halifax, Nova Scotia, Canada (Received 16 April 1974) SUMMARY 1. Orientational differences in visual resolution were measured at a number of different luminance levels on two subjects with high astigmatism that had remained optically uncorrected until the age of 10. Because of their astigmatism both of these subjects see vertical contours more clearly than horizontal contours with the unaided eye. 2. The measurements were made using sinusoidal gratings generated on the face of an oscilloscope with the refractive error carefully corrected with lenses and with the gratings viewed through 3 mm artificial pupils. 3. Visual resolution was found to be much better for vertical than for horizontal gratings for both these subjects under these conditions. The difference between the contrast sensitivities for vertical and horizontal gratings was even evident with gratings having spatial frequencies as low as 1 c/deg, but became progressively more pronounced at higher spatial frequencies. In one of the subjects the visual acuity (the cut-off spatial frequency) for horizontal gratings was more than 3/4 of an octave lower than that for vertical gratings. 4. This is very different from the results obtained from normal subjects who typically show only a slight reduction in contrast sensitivitiy for oblique gratings but resolve vertical and horizontal gratings equally well. 5. The quantitative differences between the contrast sensitivities for vertical and horizontal gratings of both high and low spatial frequencies cannot be accounted for by either errors of focus in one meridian or by the presence of meridional aniseikonia. 6. In order to completely eliminate any optical explanations for these findings measurements of contrast sensitivity were made using sinusoidal interference fringes formed directly on the retina, thereby bypassing the eye's optics. Since the orientational differences in resolution persisted with this method it must be concluded that they are of neural origin. 7. By analogy with the effects on cortical physiology that follow early selective visual deprivation in cats and monkeys, it is argued that these
740 D. E. MITCHELL AND F. WILKINSON orientational differences in resolution are a consequence of changes induced in the neural organization of the astigmat's visual system by the distorted visual input provided by the uncorrected astigmatism early in life. It is furthermore argued that the smaller orientational differences in resolution observed in normal eyes might similarly be induced by certain asymmetries in the early visual input. INTRODUCTION Recent neurophysiological studies of the effects of early selective visual deprivation on the properties of neurones in the cat and monkey visual cortex provide a fresh insight into the way in which our ability to perceive the visual world develops in early life. A number of experiments of this nature have established that the visual response characteristics of neurones in the cat and monkey visual cortex are profoundly influenced by the nature of the animal's early visual experience. This first became evident from studies which showed that the proportion of neurones that could be influenced through both eyes could be reduced to near zero by depriving kittens of congruent binocular visual input early in life (Wiesel & Hubel, 1963, 1965a; Hubel & Wiesel, 1965). Later, experiments by Hirsch & Spinelli (1970) and by Blakemore & Cooper (1970) who restricted the early visual input of each eye of young kittens to stripes of a single orientation, showed that the orientation specificity of cortical neurones was also influenced by the animal's early visual environment. Subsequently, other properties of cortical neurones such as their disparity specificity (Shlaer, 1971) and directional selectivity (M. Cynader, N. Berman & A. Hein, in preparation) have also been shown to be subject to early environmental modification. There appears to be only a short period early in the animal's life during which visual cortical neurones are susceptible to environmental modification. In the cat this period of susceptibility begins during the fourth week and ends at about 3 months of age (Hubel & Wiesel, 1970; Blakemore, 1974). The susceptible period begins at birth in the monkey, but ends at a similar age (von Noorden, 1973; Baker, Grigg & von Noorden, 1974). As the notion of a critical period implies, the consequences of early selective exposure are permanent: restoration of normal visual input after the age of 3 months fails to reverse the effects of the early abnormal exposure (Wiesel & Hubel, 1965b; Pettigrew, Olson & Hirsch, 1973). The physiological changes that are produced by early selective visual deprivation are accompanied by concordant alterations in the animal's perceptual abilities. The changes in cortical physiology induced by early deprivation of pattern vision to one eye have been shown to be accom-
VISUAL RESOLUTION IN ASTIGMATS 741 panied by a severe reduction in the visual acuity of the deprived eye (Dews & Wiesel, 1970; von Noorden, 1973), while early selective exposure to stripes of a single orientation results in reduced visual acuity for contours orthogonal to those that were present in the animal's early visual environment (Muir & Mitchell, 1973). Recent studies of humans with high astigmatism, whose early visual experience was consequently abnormal, suggests that the neural organization of the human visual system may also be influenced by its early visual input (Freeman, Mitchell & Millodot, 1972; Mitchell, Freeman, Millodot & Haegerstrom, 1973; Freeman & Thibos, 1973; Fiorentini & Maffei, 1973). Astigmatism is an optical defect in which one or more of the refractive surfaces of the eye have a toroidal shape. Consequently, the refractive power of the eye is not the same in all meridians with the meridians of maximum and minimum power, which are referred to as the principal meridians (or axes) of the astigmatism, perpendicular to each other. A distant point object is imaged by such an optical system as two focal lines parallel to the axes of the astigmatism. As a result of this an astigmat habitually sees contours parallel to one of the principal meridians more clearly than contours parallel to the other. Available clinical evidence suggests that in cases of high astigmatism the condition was present from birth so that the visual experience of these subjects would be abnormal as described until the optical error was corrected (Hirsch, 1963; Duke- Elder, 1969; Duke-Elder & Abrams, 1970). But even after full correction of the optical error many adult astigmats continue to show a substantially reduced acuity, a 'meridional amblyopia' (Martin, 1890), for contours that were most defocused before correction (Freeman et al. 1972; Mitchell et al. 1973) which suggests that the visual system of these subjects may have been modified by the anomalous early visual input they received. In this paper we report on an extensive investigation of the orientation differences in resolution exhibited by two optically corrected astigmats for gratings of both high and low contrast. The results are compared to similar measurements made earlier in normal observers (Campbell, Kulikowski & Levinson, 1966). The results suggest that the anisotropy observed in astigmatic observers is a simple exaggeration of the smaller anisotropy observed in normal subjects (Taylor, 1963). METHODS Sinusoidal gratings were generated on the face of a Telequipment D83 oscilloscope with a P31 phosphor by conventional means (outlined in Campbell & Green, 1965). The accelerating potential of this oscilloscope was sufficiently high (15 ky) to permit experiments to be performed at a luminance of 170 cd/m2. By scanning the
742 D. E. MITCHELL AND F. WILKINSON gratings formed on the oscilloscope face at this high luminance with a photomultiplier tube having a narrow slit aperture, it was found that the contrast of the gratings was directly proportional to the voltage of the sinusoidal signal applied to the Z axis from zero to 0*6 contrast. The contrast of the gratings was standardized before each experimental session by placing a neutral density filter of known transmission over the brighter half-cycle of a low frequency square-wave grating formed on the screen. The voltage of the signal to the Z axis was adjusted until a brightness match was obtained by eye with the darker half-cycle of the grating. From knowledge of the transmission of the filter the contrast of the grating at this, and any other voltage could be calculated. In order to permit the orientation of the gratings to be altered the oscilloscope was placed on its back (so that its face pointed upwards) on a large circular turntable. The subject, positioned by means of a headrest, viewed the oscilloscope face from a distance of 162 cm by means of a tilted front surfaced mirror placed directly above the oscilloscope. The orientation of the grating was altered by simply rotating the turntable underneath the mirror. A circular aperture placed on the oscilloscope face masked the field to 20 subtense. The gratings were flashed on and off at 2 Hz since both subjects found it easier to make threshold settings under these conditions. This always occurred before flicker was perceived (Kulikowski & Tolhurst, 1973). Contrast thresholds were set by having an experimenter adjust the contrast with a logarithmic attenuator in 0-05 log unit steps beginning from a value below threshold until the subject reported that he could see the gratings. The subjects were able to determine their cut-off spatial frequencies by adjusting a knob which altered the frequency of the signal applied to the Z-axis of the oscilloscope. Interference fringe-8 Sinusoidal gratings were also generated directly on the retina by focusing two beams of monochromatic light of 632-8 nm from a neon-helium laser (Metrologic Model 310) in the plane of the subject's pupil by a method described in detail elsewhere (Mitchell, Freeman & Westheimer, 1967). The fringes had a retinal illuminance of about 600 td and filled a circular field subtending 3 degrees. Subjects Data were obtained from two a-stigmatic subjects, D.E.M. (one of us) and A.M., who had both remained optically uncorrected until 10 years of age. A.M., aged 36 years, had high astigmatism in both eyes (R. - 0-25D sph, cyl - 4-5OD axis 200; L. cyl - 4-OOD axis 164') and showed markedly reduced acuity for horizontal gratings. After initial correction at 10 years of age, spectacleswereworn continuously. D.E.M., aged 31 years, was less astigmatic (R. and L. + 1-50D sph, cyl - 2*75D axis 180') and also showed reduced acuity for horizontal gratings although the deficit was not as pronounced as with A.M. Spectacle corrections have been worn continuously since initial correction at 10 years of age. Nomenclature Standard ophthalmic nomenclature (Emsley, 1955) has been used to specify the axis of the astigmatism. From the point of view of an observer viewing the subject's face, angles increase anti-clockwise from 00 at 3 o'clock to 1800 at 9 o'clock. The same nomenclature is commonly used to describe the orientation of gratings in space. As a consequence, the orientation of contours that the subject reports as clearest and most blurred with the unaided eye (which are parallel to the axes of the astigmatism) are the mirror images of the axes specified according to ophthalmic nomenclature.
VISUAL RESOLUTION IN ASTIGMATS74 743 R~ESULTS Initially a measurement was made to determine the highest spatial frequency at which gratings of different orientations could be resolved. The subjects wore their optimum optical corrections for these measurements and observed the gratings through 3 mm artificial pupils. The results for the two eyes of A.M. and the right eye of D.E.M. are shown in Fig. 1la and b respectively. The points represent the mean of six settings made at each orientation; the s.e. of the mean of these settings rarely exceeded 0-5 cfdeg. In confirmation of earlier findings (Mitchell et al. 1973), both subjects showed substantial differences between their acuities for gratings -parallel to the major axes of the astigmatism even after careful correction of the optical error. In the case of A.M.'s right eye, resolution was best for gratings oriented at 750 (28.65 c/deg) and was worst for gratings perpendicular to this (17.15 c/deg). With the other eye, acuity was best at 1050 (24.25 c/deg) and was again very depressed for gratings of the orthogonal orientation (14-67 c/deg). For ID.E.M. acuity was best when the gratings were vertical (37.4 c/deg), was depressed when they were horizontal (32.5 c/deg), and was worst when they were obliquely oriented. The reduced acuity that A.M. demonstrated for near-horizontal gratings, which amounted to 0-80 of an octave for the right eye and 0-79 of an octave for the left, was only slightly less than the severest deficit (one octave) observed among the subjects of a previous investigation (Mitchell et al. 1973). The depressed acuity that these astigmatic subjects show for horizontal gratings cannot be improved with lenses. This can be seen from Fig. 2 which shows the effect of adding lenses of various powers to the refractive correction determined by standard clinical procedures on the acuity for gratings parallel to the major axes of the astigmatism. The eye was homatropinized for the measurements on D.E.M. in order to eliminate the effect of fluctuations of accommodation. It can be seen that the addition of lenses of any power resulted in a reduction in acuity for both grating orientations. Perhaps the strongest argument against optical explanations for the meridional variations in acuity is the fact that they persist when the measurement is made with sinusoidal interference fringes formed directly on the retina (Westheimer, 1960; Campbell & Green, 1965). This technique effectively bypasses the optics of the eye since errors of focus of the eye in any meridian cannot alter either the contrast or the spatial frequency of the interference fringes formed on the retina (Campbell et al. 1966). In Fig. 3 are shown the results of measurements made on the right eye of D.E.M. of the changes with orientation of the contrast sensitivity for
744 D. E. MITCHELL AND F. WILKINSON ba 0 '~4 00 we)'40 00 450 900 1350 1800 Orientation Fig. 1. Orientation differences in visual resolution measured with sinusoidal gratings on two optically corrected astigmats. Each point represents the mean of six settings. A, open and filled circles indicate the results for the right and left eyes of A.M. respectively. B, the results for the right eye of D.E.M. The gratings had a contrast of 0-6, a mean luminance of 170 cd/rn2 and were viewed through 3 mm artificial pupils. The open and filled arrows indicate the orientation of distant contours that are most defocused by the astigmatic right and left eyes respectively.-= 00,,/=450, j=900, \= 135'.
VISUAL RESOLUTION IN ASTIGMATS 745 sinusoidal interference fringes having a spatial frequency of 22 c/deg. Also shown are the results of similar measurements made with sinusoidal gratings of the same spatial frequency formed on an oscilloscope and viewed through a 3 mm artificial pupil. The retinal illuminance of the gratings was adjusted to be the same, about 600 td, for both sets of measurements. Although the contrast sensitivities for sinusoidal interference A 301F bo 0 lua U e C Cr 0" - a. Mf I I IU I I I I B 40r- 201-201- (r I 0-- coo a, 10 I I I I I I I -1-50 -100-0*50 0 0*50 1-00 1*50 Added lens power (dioptres) Fig. 2. The effect on visual resolution of adding lenses of various powers to the spectable corrections of the two astigmats of Fig. 1. A, the effects on vertical (0) and horizontal (0) gratings respectively for the homatroponized right eye of subject D.E.M. B, the results of similar measurements made with gratings oriented at 750 (0) and 1650 (0) on the right eye of A.M. The gratings, which had a mean luminance of 170 cd/m2, were viewed through 3 mm artificial pupils. The data for D.E.M. were obtained with gratings of 0-6 contrast while those for A.M. were obtained at a contrast of only 0*38.
746 746 D. E. MITCHELL AND F. WILKINSON fringes were higher than those for gratings imaged by the eye's optics, the general shape of the curves was very similar. Thus it can be concluded that these resolution differences are not optical in origin. Aside from an uncorrected error of focus in one meridian, there is one other possible optical explanation for the meridional amblyopia. The presence of any meridional variation in the magnification of the image on the retina could result in an orientational difference in resolution even if the image was exactly focused in every meridian. This condition, called 50 40 ~30--. /~~~~~~~~~C o0 ~20 U / 10~~~~ 00 450 901350 1800 Orientation Fig. 3. Orientation differences in thc contrast sensitivity for sinusoidal gratings generated on an oscilloscope (0) and for sinusoidal. interference fringes generated directly on the retina by a laser (0) measured on subject D.E.M. The spatial frequency (22 c/deg) and the luminance (600 td) were adjusted to be the same for both sets of measurements. aniseikonia, is encountered in some astigmats and can result in dramatic perceptual distortions particularly in cases where the magnification in one meridian is different in the two eyes (Ogle, 1964). The lowered resolution that our two subjects show for horizontal gratings could be explained quite simply by the presence of a meridional aniseikonia which effectively expanded the spatial frequency scale for vertical as compared to horizontal gratings by factors of 1-76 and 1-28 for A.M. and ID.E.M. respectively. However, the measurements made with interference fringes (Fig. 3) effectively eliminates this as a possible explanation since this technique bypasses the effects of meridional aniseikonia (Campbell et al. 1966).
VISUAL RESOLUTION IN ASTIGMATS 747 The large meridional variations in resolution that these subjects show with high contrast gratings are also seen with gratings of low contrast. This is apparent from Fig. 4 which shows the results of measurements made on the left eye of A.M. and on D.E.M. of the orientational changes in the contrast sensitivity (the reciprocal of the threshold contrast) for._._ I10._ r 4-68 c/deg 1-17 c/deg 16-3 c/deg 18-7 c/deg C VI 0 1-17 c/deg 7 0 c/deg 11-7 c/deg 00 450 900 1350 1800 Orientation Fig. 4. Contrast sensitivity for sinusoidal gratings as a function of orientation measured at several different spatial frequencies. The gratings had a luminance of 170 cd/m2 and were viewed through 3 mm artificial pupils. A, results obtained from the right eye of D.E.M. B, the results of similar measurements made on the left eye of A.M. sinusoidal gratings of certain fixed spatial frequencies. The size of the field was increased to 5.40 for the measurements with gratings of the lowest spatial frequency (1.17 c/deg) in order to increase the number of cycles of the grating that were visible. A small but statistically significant difference (largest P = 0-005) existed between the contrast sensitivities
748 D. E. MITCHELL AND F. WILKINSON._._ 0 500 r 200 100 50 20 0 W 10 C 5 2 ;D S ~0 \ 00 170 cd/r2 I I I I I I 1 I., 'I 10 cd/m2 ' 0 I I I VI I 1-7 cd/m2 0_Ak 00 l I * Xl I 10 20 30 40 10 20 0 10 20 Spatial frequency (c/deg) Fig. 5. Contrast sensitivity as a function of spatial frequency for sinusoidal gratings with orientations of 750 (@) and 1650 (0) measured on the right eye of A.M. at three luminance levels. 1000 _ 500 200 G170 cd/m2 \0 20 7 cd/m2 lo0 0 la'i 1 cd/rn2 >4 \ N C 20 - U 10 0\ 5 C\~~~~~ 2-0 10 20 30 40 50 Spatial frequency (c/deg) Fig. 6. Contrast sensitivity as a function of spatial frequency for vertical (*, *, A) and horizontal (0, E, A) sinusoidal gratings at three luminance levels measured on the right eye of D.E.M.
VISUAL RESOLUTION IN ASTIGMATS 749 for gratings parallel to the principal axes of the astigmatism even when the spatial frequency of the gratings was as low as 1P17 c/deg. These differences became progressively larger at higher spatial frequencies. The contrast sensitivities for gratings parallel to the principal meridians of the astigmatism (900 and 1800 for D.E.M.; 750 and 1650 for the right eye of A.M.) are depicted further in Figs. 5 and 6 for a wide range of spatial frequencies at three different luminance levels. The results have been plotted on semilogarithmic co-ordinates since the data can be approximated by straight lines for spatial frequencies above about 5 c/deg when plotted in this way (Campbell & Green, 1965). The data show both the well known low and high frequency attenuation. At the highest luminance level (170 cd/m2) the sensitivity is highest at 4-5 and 6 c/deg for A.M. and D.E.M. respectively and is lower for gratings of either higher or lower spatial frequency. As has been previously noted (Van Nes & Bouman, 1967), the peak sensitivity shifts to lower spatial frequencies with reduction in the mean luminance. This shift towards lower spatial frequencies at low luminance levels appears to be greater in the amblyopic than in the normal meridian for these astigmatic subjects. The slope of the line that best fit the data for spatial frequencies above about 5 c/deg changes with the orientation of the gratings and is greatest for gratings that are horizontal (D.E.M.) or at 1650 for these particular subjects. At lower luminances the difference between the contrast sensitivities for gratings parallel to the principal meridians of the astigmatism becomes smaller as does the difference between the slope of the lines drawn through the data for gratings of these orientations. Comparison with normal observers Fig. 7 shows the results of measurements of the contrast sensitivities for vertical, horizontal, and oblique gratings made on a normal observer under identical conditions to those employed for the astigmatic subjects. In confirmation of the usual findings (Campbell et al. 1966; Mitchell et al. 1967) sensitivity was greater when the gratings were either vertical or horizontal than when they were oblique. This particular subject. however, could resolve oblique gratings at 45 somewhat better than those at 1350. As Campbell et al. (1966) noted, the data for oblique gratings are fitted by lines with a steeper gradient than that which best fits the results obtained with horizontal or vertical gratings. The ratio of the slopes for oblique (at 1350) versus horizontal and vertical gratings was 1-12 in this subject, which is somewhat smaller than the average value of 1 22 that Campbell et al. (1966) report for their three subjects. Comparison of the data shown is Fig. 7 with that of the astigmats in
750 D. E. MITCHELL AND F. WILKINSON Figs. 5 and 6 indicates that the astigmatic subjects possessed poorer resolving power than normal even in their best meridian. This was particularly evident for A.M. for whom the estimated cut-off spatial frequency for nearly vertical gratings was more than half an octave lower than that of the normal subject. 1000 Soo 200 100 C 4._ 0 I, C 0 U 50 20 10 5 2 I 10 20 30 40 50 Spatial frequency (c/deg) Fig. 7. Contrast sensitivity as a function of spatial frequency for sinusoidal gratings of four different orientations. *, vertical (900); 0O horizontal (00); A, oblique (450); A, oblique (1350). The gratings had a mean luminance of 170 cd/m2 and were viewed through a 3 mm artificial pupil. Campbell et al. (1966) found that above about 7 c/deg the contrast sensitivities for normal observers could be fitted by the equation S(f) = A exp (-akf), (1) where f is the grating spatial frequency, A is a coefficient dependent upon dioptric factors and luminance, a is the slope for vertical gratings and kl a correction coefficient dependent upon the orientation, a, of the grating. The value of a for vertical or horizontal gratings was typically 0-13 deg/cycle. Inspection of Figs. 5 and 6 show that above 5 c/deg eqn. (1) also describes the data of the astigmatic subjects. At 170 cd/m2 the value of a
VISUAL RESOLUTION IN ASTIGMATS 751 for vertical or nearly vertical gratings was 0 14 deg/cycle for D.E.M. and 0*15 deg/cycle for A.M. which agrees closely with the value obtained (0.13) on the normal observer for horizontal or vertical gratings. The change in the slope with alteration in the grating orientation was, however, very much larger than that observed in normal observers. Whereas the ratio of the slopes for oblique versus horizontal or vertical gratings was 1*12 for the normal subject of Fig. 7, the ratio of the slopes for gratings parallel to the major axes of the astigmatism for A.M. and D.E.M. at 170 cd/m2 were 1*76 and 1*28 respectively. D.E.M. 0 100 7 - -,0 A.M. 10 10 1 2 5 10 20 50 1 2 5 10 20.50 Spatial frequency (c/deg) Fig. 8. Contrast sensitivity as a function of spatial frequency replotted from Figs. 5 and 6 (170 cd/m2) on to logarithmic scales. Results are shown for two grating orientations, vertical (@) and horizontal (Q) for D.E.M. and 75 (@) and 165 (0) for A.M. A continuous line has been fitted by eye through the data for vertical or near vertical gratings. The interrupted curve is this same curve displaced to the left by a factor of 1-76 for A.M. and 1-28 for D.E.M. The results of Figs. 5 and 6 provide additional arguments against optical explanations for the meridional amblyopia. The fact that the contrast sensitivities for gratings parallel to the principal meridians of the astigmatism fall on lines that differ in slope rules out the possibility that this difference results from an error of focus in one meridian since the effect of focusing errors is to uniformly decrease the contrast sensitivity for gratings with spatial frequencies above 5 c/deg (Campbell & Green, 1965; Campbell et at. 1966; D. E. Mitchell & F. Wilkinson, unpublished observations). TheX data of Figs. 5 and 6 also argue against meridional aniseikonia as a cause of the meridional amblyopia. Since the effect of any
752 D. E. MITCHELL AND F. WILKINSON meridional aniseikonia would be to magnify all spatial frequencies in one meridian equally, the data for gratings parallel to the axes of the astigmatism should be displaced from each other by factors of 1*76 and 1*28 for A.M. and D.E.M. respectively. In order to better illustrate the results for low spatial frequencies the data of Figs. 5 and 6 have been replotted in Fig. 8 on logarithmic co-ordinates. A continuous curve has been fitted by eye through the data for vertical (or nearly vertical) gratings. Displacement of this line to the left (interrupted lines) by factors of 1-76 (A.M.) and 1*28 (D.E.M.) provides a good fit for high spatial frequencies, but below 5 to 10 c/deg the fit is poor. Relation to early visual experience Fig. 9 shows the disposition of the focal lines relative to the retina of A.M. and D.E.M. when the eye is optically uncorrected and accommodation is relaxed. In the case of A.M. (Fig. 9a), who has simple myopic astigmatism, the horizontal focal line is the most defocused when the eye is unaccommodated and thus distant horizontal contours will always be seen less clearly than vertical ones by this subject in all states of A 8 C Fig. 9. The disposition of the two focal lines with respect to the retina for two astigmats. A, the position of the focal lines for subject A.M. when accommodation is relaxed. The nearly vertical focal line is coincident with the retina in this situation. B and C, the position of the focal lines for D.E.M. When accommodation is fully relaxed (B) the two focal lines almost exactly straddle the retina. However, by accommodating about 1S5D (C) the vertical focal can be brought into coincidence with the retina.
VISUAL RESOLUTION IN ASTIGMATS 753 accommodation of the eye. However, this is not always true for D.E.M. who has mixed astigmatism. When the eye is unaccommodated (Fig. 9b), the two focal lines almost exactly straddle the retina so that under these conditions both horizontal and vertical contours would appear equally blurred. But as is shown in Fig. 9c, it is possible for this subject to clearly image distant vertical contours, at the expense of horizontal contours, by slightly increasing the state of accommodation. In fact, both direct measurements of the accommodative responses (Mitchell et al. 1973; R. Freeman, in preparation), as well as the subjective reports this subject makes when viewing contours of different orientations, show that he does indeed preferentially accommodate for vertical contours in the visual field in this way. As a consequence, distant vertical contours are habitually seen more clearly than horizontal ones with the unaided eye. The results for A.M. shown in Fig. la provides an excellent example of the close agreement between vision with and without corrective lenses. The arrows at the top of Fig. la indicate the orientation of distant contours that are most defocused by the astigmatic optics; clearly these coincide very closely with the grating orientation for which acuity is worst. DISCUSSION There are two plausible explanations for the abnormal meridional variations in acuity observed in these astigmats. First, it is possible that the two conditions arise entirely independently of each other, presumably in response to a common causal factor. However, the close correlation between the optical defect and the neural anomalies demanded by this explanation is most unlikely to arise through two such independent processes; hence this must be considered to be an extremely remote possibility. On the other hand, there is much evidence in favour of the alternative explanation which would propose that the meridional amblyopia arose as a consequence of changes induced in the neural organization of the visual cortex by the abnormal early visual input (Blakemore & Cooper, 1970; Mitchell et al. 1973; Freeman & Pettigrew, 1973). While the deficits observed in these astigmats are qualitatively concordant with the nature of their early visual input, they nevertheless differ somewhat from the changes in contrast sensitivity that result in an adult when the retinal image is defocused. For example, it is noteworthy that both subjects showed a statistically significant deficit for horizontal gratings having spatial frequencies as low as 1-17 c/deg. But errors of focus of the magnitude expected in these two subjects have negligible effects on the visibility of gratings of this spatial frequency in an adult and would be unlikely to do so in an infant (Campbell & Green, 1965). From this it
754 D. E. MITCHELL AND F. WILKINSON would appear that the perceptual deficits extend further than the immediate effects of the original error of focus. This would suggest that the neurones that detect gratings of low spatial frequency were sufficiently broadly tuned for spatial frequency early in life in order to have been influenced during their development by the asymmetrical visual input that was present at higher spatial frequencies. Certainly neurones in the visual cortex of kittens during the first month of life exhibit far less selectivity than in the adult (Hubel & Wiesel, 1963; Barlow & Pettigrew, 1971; Pettigrew, 1972). In view of the poor visual spatial resolution of the human infant in the first 2 months of life (Atkinson, Braddick & Braddick, 1974) it is quite likely that this is also true of neurones in the new-born human visual cortex. Finally, the results of Figs. 5 and 6 show that the depressed sensitivity that these two astigmats show for horizontal gratings is a simple exaggeration of the reduced sensitivity that normal observers show for oblique gratings which suggests that the two conditions may be of similar origin. This adds some support for the suggestion (Annis & Frost, 1973) that the reduced acuity that normal subjects show for oblique stimuli develops early in life as a consequence of a bias induced in the orientation preference of cortical neurones by the relatively low frequency of occurrence of oblique contours in most urbanized cultures. The research for this paper was supported in part by the Defence Research Board of Canada, Grant number 9401-58, and by the Medical Research Council of Canada (MA 5027). REFERENCES ANNIs, R. C. & FROST, B. (1973). Human visual ecology and orientation anisotropies in acuity. Science, N.Y. 182, 729-731. ATKINSON, J., BRADDICK, 0. & BRADDICK, F. (1974). Acuity and contrast sensitivity of infant vision. Nature, Lond. 247, 403-404. BAKER, F. H., GRIGG, P. & von NOORDEN, G. K. (1974). Effects of visual deprivation and strabismus on the response of neurons in the visual cortex of the monkey, including studies on the striate and prestriate cortex in the normal animal. Brain Re8. 66, 185-208. BARLow, H. B. & PETTIGREW, J. D. (1971). Lack of specificity of neurones in the visual cortex of young kittens. J. Physiol. 218, 98-100P. BLAKEMORE, C. (1974). Developmental factors in the formation of feature extracting neurons. In The Neurosciences, Third Study Program, ed. WOORDEN, F. G. & ScMITrr, F. 0. Cambridge, Mass.: M.I.T. Press. BLAKEMORE, C. & COOPER, G. F. (1970). Development of the brain depends on the visual environment. Nature, Lond. 228, 477-478. CAMPBELL, F. W. & GREEN, D. G. (1965). Optical and retinal factors affecting visual resolution. J. Physiol. 181, 576-593. CAMPBELL, F. W., KULiKOWSKI, J. J. & LEVINSON, J. (1966). The effect of orientation on the visual resolution of gratings. J. Physiol. 187, 427-436.
VISUAL RESOLUTION IN ASTIGMATS 755 DEWS, P. B. & WIESEL, T. N. (1970). Consequences of monocular deprivation on visual behaviour in kittens. J. Physiol. 206, 437-455. DUKE-ELDER, W. S. (1969). The Practice of Refraction. St Louis: Mosby. DUKE-ELDER, W. S. & ABRAMS, D. (1970). Ophthalmic optics and refraction. In System of Ophthalmology, vol. v, ed. DUKE-ELDER, W. S. London: Henry Kimpton. EMsLEY, H. H. (1955). Visual Optics, vol. I, 5th edn. London: Hatton Press. FIORENTINI, A. & MAFFEI, L. (1973). Evoked potentials in astigmatic subjects. Vision Res. 13, 1781-1783. FREEMAN, R. D., MITCHELL, D. E. & MILLODOT, M. ( 1972). A neural effect of partial visual deprivation in humans. Science, N.Y. 175, 1384-1386. FREEMAN, R. D. & PETTIGREW, J. D. (1973). Alteration of visual cortex from environmental asymmetries. Nature, Lond. 246, 359-360. FREEMAN, R. D. & THIBOS, L. N. (1973). Electrophysiological evidence that abnormal visual experience can modify the human brain. Science, N.Y. 180, 876-878. HIRSCH, H. V. B. & SPINELLI, D. N. (1970). Visual experience modifies distribution of horizontally and vertically oriented receptive fields in cats. Science, N.Y. 168, 869-871. HIRSCH, M. (1963). The refraction of children. In Vision of Children, ed. HIRSCH, M. & WICKS, R. Philadelphia: Chilton. HUBEL, D. H. & WIESEL, T. N. (1963). Receptive fields of cells in striate cortex of very young, visually inexperienced kittens. J. Neurophysiol. 26, 994-1002. HUBEL, D. H. & WIESEL, T. N. (1965). Binocular interaction in striate cortex of kittens reared with artificial squint. J. Neurophysiol. 28, 1041-1059. HUBEL, D. H. & WIESEL, T. N. (1970). The period of susceptibility to the physiological effects of unilateral eye closure in kittens. J. Physiol. 206, 419-436. KULIKOWSKI, J. J. & TOLHURST, D. J. (1973). Psychophysical evidence for sustained and transient detectors in human vision. J. Physiol. 232, 149-162. MARTIN, G. (1890). Theorie et clinique de l'amblyopie astigmatique. Annals Oculist. 104, 101-138. MITCHELL, D. E., FREEMAN, R. D. & WESTHEIMER, G. (1967). Effect of orientation on the modulation sensitivity for interference fringes on the retina. J. opt. Soc. Am. 57, 246-249. MITCHELL, D. E., FREEMAN, R. D., MILLODOT, M. & HAEGERSTROM, G. (1973). Meridional amblyopia: Evidence for modification of the human visual system by early visual experience. Vision Res. 13, 535-558. MUIR, D. W. & MITCHELL, D. E. (1973). Visual resolution and experience: acuity deficits in cats following early selective visual deprivation. Science, N.Y. 180, 420-422. OGLE, K. N. (1964). Researches in Binocular Vision. New York: Hafner. PETTIGREW, J. D. (1972). The importance of early visual experience for neurons of the developing geniculo-striate system. Invest. Ophthal. 11, 386-394. PETTIGREW, J. D., OLSON, C. & HIRSCH, H. V. B. (1973). Cortical effect of selective visual experience: degeneration or reorganization? Brain Res. 51, 345-351. SHLAER, S. (1971). Shift in binocular disparity causes compensatory change in the cortical structure of kittens. Science, N.Y. 173, 638-641. TAYLOR, M. M. (1963). Visual discrimination and orientation. J. opt. Soc. Am. 53, 763-765. VAN NES, F. L. & BOUMAN, M. A. (1967). Spatial modulation transfer in the human eye. J. opt. Soc. Am. 57, 401-406. von NOORDEN, G. K. (1973). Experimental amblyopia in monkeys. Further behavioral observation and clinical correlations. Invest. Ophthal. 12, 721-726.
756 D. E. MITCHELL AND F. WILKINSON WESTIEIMER, G. (1960). Modulation thresholds for sinusoidal light distributions on the retina. J. Physiol. 152, 67-74. WIESEL, T. N. & HUBEL, D. H. (1963). Single-cell responses in striate cortex of kittens deprived of vision on one eye. J. Neurophysiol. 26, 1003-1017. WIESEL, T. N. & HUBEL, D. H. (1965a). Comparison of the effects of unilateral and bilateral eye closure on cortical unit responses in kittens. J. Neurophysiol. 28, 1029-1040. WIESEL, T. N. & HUBEL, D. H. (1965b). Extent of recovery from the effects of visual deprivation in kittens. J. Neurophysiol. 28, 1060-1072.