The Relationship Between Binocular Rivalry and Strabismic Suppression

Earl L. Smith III, Dennis M. Levi, Rurh E. Manny, Ronald 5. Harwerrh, and Janis M. Whire

Increment-threshold spectral sensitivity functions were determined for normal observers during binocular rivalry and for esotropic observers during strabismic suppression and under viewing conditions that normally induce binocular rivalry. Depending on the spatial and temporal properties of the test stimulus, the normal observers exhibited a wavelength-specific loss in sensitivity during the suppression phase of rivalry, which suggests that binocular rivalry differentially attenuates the sensitivity of the chromatic mechanisms relative to the luminance mechanisms. In contrast, regardless of the test stimulus dimensions, the esotropic observers did not manifest a wavelength-specific loss in sensitivity either during strabismic suppression or under conditions that normally induce binocular rivalry. The different patterns of suppression shown by the normal and esotropic subjects suggest that strabismic observers do not demonstrate normal binocular rivalry, and that strabismic suppression and normal binocular rivalry suppression are mediated by different neural mechanisms. Invest Ophthalmol Vis Sci 26:80-87, 1985

The phenomenon of suppression is characterized Because both binocular rivalry suppression and by an inability to perceive normally visible objects in strabismic suppression result in the functional loss of all or part of the visual field of one eye. Suppression visual information from one eye during binocular occurs under binocular viewing conditions and is viewing conditions, it has been hypothesized in the considered to represent an interocular inhibitory pro- classic1 and more recent literature3'4 that binocular cess that prevents visual information from the "sup- rivalry suppression and strabismic suppression are pressed" eye from reaching the threshold for conscious related. If binocular rivalry suppression and strabismic perception.1 Suppression phenomena occur in indi- suppression are related, it is reasonable to assume viduals with normal binocular vision as well as in that common neural mechanisms may underlie both individuals with anomalous binocular vision. For forms of suppression. To date, little is known about example, when dissimilar targets that cannot be fused the nature of the mechanisms that are responsible for into a single percept are presented simultaneously to binocular rivalry and strabismic suppression. Mea- corresponding retinal areas in the two eyes of an surements of the changes in visual sensitivity produced observer with normal binocular vision, the observer by binocular rivalry suppression and strabismic experiences an unstable alternating suppression of suppression would provide a way to characterize and information from each eye, a phenomenon called compare the underlying mechanisms responsible for binocular rivalry.2 Suppression is also commonly these two types of suppression. In a previous study,5 exhibited by strabismic observers. Many individuals we measured the spectral sensitivity of the eyes of with strabismus of early onset reduce or eliminate normal observers during binocular rivalry and found the distressing sensory aspects of their eye misalign- a wavelength-specific loss in sensitivity, which sug- ment (diplopia and confusion) by suppressing the gested that binocular rivalry suppression differentially conflicting visual information from specific regions attenuates the sensitivity of the chromatic mechanisms of the deviated, nonfixating eye.13 relative to the luminance mechanisms. Because of the distinctive, wavelength-specific loss of visual sen- sitivity during binocular rivalry, spectral sensitivity From the University of Houston—University Park, College of measurements appear to provide a reliable means for Optometry, Houston, Texas. characterizing and comparing the nature of the inhib- Supported by National Eye Research grants R01-EY03611, R01- itory states associated with binocular rivalry and EY01139, and R01-EY01728 and BRSG grant SO7-RR07147. strabismic suppression. Therefore, in the present study, Submitted for publication: March 14, 1983. Reprint requests: Earl L. Smith III, OD, PhD, University of we compared the spectral sensitivity functions ob- Houston, University Park, College of Optometry, Houston, TX tained for normal observers during binocular rivalry 77004. with the functions obtained for esotropic observers

80

Downloaded from iovs.arvojournals.org on 09/28/2021 No. 1 BINOCULAR RIVALRY AND STRABISMIC SUPPRESSION / Smirh III er ol. 81

Table 1. Visual characteristics of the strabismic observers

Fixation status of Observer deviating eye no. Age/sex Eye Refractive status Acuity (visuoscopy) Binocular status

1 22/F OD* piano 20/40 Unsteady central Congenital esotropia. OSf piano 20/15 Strabismus surgery at age 7. Presently constant 4A right esotropia with 2A hypotropia

2 23/F OD +4.50 - 0.75 X 100 20/22 Unsteady central with Constant 2A left esotropia OS +4.75 - 1.00 X 75 20/30 nasal drift 0.5 deg 3 26/M OD +0.75 - 0.25 X 70 20/20 Unsteady 1° nasal Constant 8A left esotropia OS +6.50 - 0.50 X 23 20/60 Eccentric fixation 4 25/M OD -6.75 DSt 20/30 0.5 deg nasal eccentric Constant 3A right esotropia OS -6.25 DS 20/20 fixation 5 36/M OD -1.25 - 1.25 X 52 20/25 Unsteady central Alternating esotropia noted OS -2.00 - 0.50 X 05 20/20 since early childhood. Strabismus surgery at age 29. Presently 15" constant right esotropia with hypotropia

• OD = right eye; fOS = left eye. i Dioptric sphere.

during strabismic suppression and under conditions Apparatus and Procedures that normally induce binocular rivalry. The results clearly demonstrate that different mechanisms are The stimuli that were used to induce binocular involved in normal binocular rivalry and strabismic rivalry and strabismic suppression were high-contrast, suppression. square-wave gratings with a fundamental spatial fre- quency of 2.8 cycles per degree and a space-averaged 2 Materials and Methods luminance of 5.9 cd/m . The gratings were surrounded by a 0.5-deg dark border to form a square pattern Subjects that subtended 5.7 deg. The grating patterns were embedded in a spatially homogenous 19-deg field, Five strabismic individuals, demonstrating constant with a luminance of 17 cd/m2. unilateral esotropia of early onset, mild amblyopia For the binocular rivalry experiments, a vertical (Davidson-Eskridge visual acuity chart6), no stereopsis grating was presented to the eye that was to be tested (Titmus Stereo Test), and normal color vision (A.O.- and a horizontal grating was presented to the other HRR Pseudoisochromatic Plates) were used as subjects eye. A mirror stereoscope was used to present the in the strabismic suppression experiments. All of the orthogonally oriented grating patterns to correspond- strabismic observers exhibited harmonious anomalous ing points in the two eyes. For the esotropic observers, retinal correspondence on the Bagolini striated glass this necessitated adjusting the stereoscope so that the test7 but normal retinal correspondence when viewing two grating patterns coincided with the observer's targets in the haploscope used in the present study. subjective angle of squint. The dark borders sur- The visual characteristics of the strabismic observers rounding the gratings provided a strong cue for fusion are summarized in Table 1. Ten individuals with and both the normal and strabismic observers reported normal visual acuity (20/20 or better in each eye), that the borders were perceived as single. However, color perception, and binocular vision were used as in agreement with previous studies of binocular ri- subjects for the binocular rivalry experiments. Two valry, the strabismic (see references 8 and 9) as well of the authors served as normal subjects, however, as the normal observers reported an alternation of all of the other observers were naive with respect to visibility between the vertical gratings seen by the the purpose of the experiment. Each observer had tested eye and the horizontal gratings seen by the clear media, normal fundi, and were appropriately other eye. The observers reported that at times the corrected for any refractive errors. entire field for one eye was suppressed completely,

Downloaded from iovs.arvojournals.org on 09/28/2021 82 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / January 1985 Vol. 26

but that usually the composite pattern consisted of tected the test probe was defined as the observer's regions of vertical bars interspersed with regions of threshold. The stimulus intensity was reduced subse- horizontal bars. quently about 0.4 to 0.5 log units and the next For the strabismic suppression experiments, verti- ascending threshold series was initiated. Three of the cally oriented gratings were presented to both eyes normal observers and all of the strabismic observers rather than the orthogonally oriented gratings utilized thresholds were measured at 20-nm intervals for to induce binocular rivalry. Vertically oriented gratings stimulus wavelengths between 420 and 680 nm. Six were used because, as Schor910 has demonstrated, threshold estimates were determined for each stimulus similar patterns that fall on corresponding points and wavelength with the order of presentation (ie, ascend- that normally stimulate sensory fusion are the optimal ing or descending) counterbalanced from session to stimuli for inducing strabismic suppression. The ver- session. The log of the reciprocal of the number of tical grating patterns contained cues for monitoring quanta associated with the geometrical mean of the suppression (small, nonpaired, horizontal black bars) six threshold estimates determined over two to three and were viewed in the mirror stereoscope adjusted experimental sessions was plotted against the wave- for the esotropic observers' subjective angle of squint. number of the test stimulus to produce spectral For individuals with normal binocular vision, the sensitivity functions for each of the viewing conditions. patterns used to induce strabismic suppression were For the normal observers, spectral sensitivity functions fused easily into a stable, single percept, and the were obtained for the dominance and suppression suppression cues for each eye were always visible. phases of binocular rivalry, and for a nonrivalrous Although the strabismic observers also reported a monocular viewing condition (the untested eye was single percept, the presence of a suppression scotoma occluded). For the esotropic observers, spectral sen- in the deviating eye was noted by the disappearance sitivity functions were obtained for the deviating eye of the corresponding suppression cues. under monocular viewing conditions (ie, nonsup- A single channel optical system consisting of a pressed control condition), when both eyes viewed heat-filtered 250 W tungsten-halogen light source, a similarly oriented patterns (ie, during strabismic monochromator (10-nm bandpass), an electronic suppression) and when both eyes viewed orthogonally shutter, neutral density wedges, a variable aperture, oriented patterns (ie, during rivalry suppression). Since and the appropriate focusing lenses was used to the shape of the spectral sensitivity function is depen- project a spectral test probe onto the center of the dent on the spatio-temporal characteristics of the test vertical gratings seen by the left eyes of the normal stimulus, these sets of spectral sensitivity functions observers. For the strabismic observers, the test probe were obtained for three different test probe sizes and was projected onto the area of the grating pattern durations (0.35 deg and 5 msec; 0.85 deg and 20 corresponding to the suppression scotoma of the msec; 1.2 deg and 50 msec). deviating eye (usually about 1 deg into the nasal Because the interpretation of the results of this field). The onset of the stimulus was controlled by study depend to some extent on the consistency of the observers who were instructed to trigger the test the wavelength-dependent sensitivity loss reported probe by depressing a button either during the appro- previously in normal observers during binocular ri- priate phase of rivalry or during strabismic suppres- valry,5 detection thresholds were measured for all 10 sion. Due to the dynamic nature of these suppression normal subjects for two stimulus wavelengths (450 phenomena, especially binocular rivalry, stimulus nm and 560 nm) during the dominance and suppres- presentations that occurred when the area of the sion phases of rivalry and under monocular viewing grating pattern corresponding to the position of the conditions. Ten threshold measurements were ob- test probe was not completely visible (eg, during tained for each observer for all of the stimulus rivalry dominance) or completely suppressed (eg, conditions. The stimulus wavelengths used in this during rivalry suppression) were repeated until the experiment were selected because our previous study5 observer was confident that the stimulus had been indicated that for the specific test probe dimensions presented at the appropriate time. used in this portion of the study (0.85 deg and 20 Detection thresholds for the test probe were deter- msec), rivalry suppression reduced the sensitivity for mined using an ascending method of limits. At the stimulus wavelengths shorter than about 480 nm to start of each ascending series, the intensity of the a much greater extent than for wavelengths longer stimulus was reduced well below the observer's detec- than 480 nm. Therefore, a comparison of the effects tion threshold. The intensity of the stimulus was of rivalry suppression on the 450 nm and 560 nm increased in 0.1 log unit steps following trials in stimuli should provide an indication of the consistency which the observer failed to detect the stimulus. The of the wavelength-dependent nature of rivalry lowest stimulus intensity at which the observer de- suppression.

Downloaded from iovs.arvojournals.org on 09/28/2021 No. 1 DINOCULAR RIVALRY AND STRADISMIC SUPPRESSION / Smirh III er ol. 83

Fig. 1. Increment-thresh- old spectral sensitivity func- tions obtained from two A. NORMAL OBSERVER B. NORMAL OBSERVER normal observers. Functions 6.0- are shown for the following three test-probe dimensions: 50 msec, 1.2 deg (circles); 20 msec, 0.85 deg (squares); and 5 msec, 0.35 deg (dia- 5.0 5.0- monds). For clarity of pre- sentation, the functions for the 50 msec, 1.2-deg test >- probe have been displaced > 4.0 upward 1.4 log units. Open »- symbols, filled symbols and 5> half-filled symbols represent z UJ data for the monocular con- (0 trol condition, rivalry UJ > 3.0-1 suppression, and rivalry dominance, respectively. UJ oc Each data point represents o the geometrical mean of six 2 2-OH 2.0- threshold estimates. The standard errors of the mean were typically less than 0.05 log units. The functions 1.0 1.0 drawn through the data for reference represent either the functions derived from the linear-subtractive interaction

model of spectral sensitivities 450 500 550 600 650 700 400 450 500 550 600 650 700 proposed by Sperling and WAVELENGTH (nm) WAVELENGTH (nm) Harwerth,i:i or functions de- 25,000 22,000 19,000 16,000 25,000 22.000 19.000 16.000 termined for normal observ- 1 WAVENUMBER (cm ) WAVENUMBER (cm") ers in a previous study using flicker methods."

Results chromatic or opponent-color mechanisms,"12 the dominance phase and monocular control functions Normal Rivalry Suppression exhibit three maxima of sensitivity at approximately Figure I illustrates increment-threshold spectral 440, 530, and 610 nm and conform to the linear, sensitivity functions determined for two of the normal subtractive-interaction model for spectral sensitivity observers during binocular rivalry and under mon- proposed by Sperling and Harwerth.13 But, in contrast, ocular viewing conditions using three different test- for the same test-probe dimensions, the suppression probe durations and field sizes. The results obtained phase data are described adequately by a function for all of the normal observers confirm and extend obtained with flicker methods (ie, a unimodal lumi- our previous investigation5 of the spectral character- nosity function with a peak at about 550 nm). The istics of binocular rivalry. The data indicate that the only departure of the suppression-phase data from shapes of the spectral sensitivity functions obtained the luminosity function occurred for the 50 msec, under the monocular viewing conditions (open sym- 1.2 deg stimulus (circles). For one of the normal bols) and the dominance phase of rivalry (half-filled observers (panel A) the suppression-phase data for symbols) are identical for all of the stimulus durations the shortest wavelengths (420 and 440 nm) appear to and field sizes. However, depending on the spatial conform to the absorption spectrum of the short- and temporal characteristics of the test probe, the wavelength sensitive cones. These differences in the spectral sensitivity functions for the suppression phase shapes of the spectral sensitivity functions for the of rivalry (filled symbols) differ markedly in shape dominance and suppression phases of rivalry are not from the functions obtained during the dominance apparent for test probe dimensions that normally bias phase of rivalry and during the monocular viewing detection toward luminance mechanisms."12 For in- condition. For example, for the larger test fields and stance, for the 5 msec, 0.35-deg test probe (diamonds), longer viewing durations (circles and squares), stim- the data obtained for all three viewing conditions ulus dimensions that normally bias detection toward conform to the shape of functions obtained by flicker

Downloaded from iovs.arvojournals.org on 09/28/2021 84 INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / January 1985 Vol. 26

BINOCULAR RIVALRY did for the 560-nm stimulus (t(9) = 6.50, P < 0.0001). NORMAL OBSERVERS This result confirms our previous investigation of the IZZ) DOMMANT PHASE spectral characteristics of binocular rivalry and further mm SUPPRESSION PHASE suggests that this wavelength-selective aspect of rivalry suppression is a characteristic of observers with normal u.o - T binocular vision. LAR)

o 0.4 - Strabismic Suppression Typical increment-threshold spectral sensitivity

Downloaded from iovs.arvojournals.org on 09/28/2021 No. 1 BINOCULAR RIVALRY AND STRADISMIC SUPPRESSION / Smirh III er ol. 85

Fig. 3. Increment-thresh- old spectral sensitivity func- A. ESOTROPIC OBSERVER (JB) B. ESOTROPIC OBSERVER (DK) tions obtained from two 6.0-1 6.0-1 esotropic observers under monocular viewing condi- tions (open symbols) and

during strabismic suppres- 5.0- 5.0- sion (filled symbols). Func- tions are shown for the fol- lowing three test-probe di- >• mensions: 50 msec, 1.2 deg (circles); 20 msec, 0.85 deg > 4.0 (squares); and 5 msec, 0.35 55 deg (diamonds). For clarity in w of presentation, the functions m for the 50 msec, 1.2-deg test > 3.0 3.0- probe have been displaced upward 1.4 log units. Each 111 OC data point represents the o geometrical mean of six 9 2.0 H threshold estimates. The standard errors of the mean were typically less than 0.05 log units. The functions drawn through the data for reference represent either the functions derived from the linear-subtractive interaction model of spectral sensitivities 400 450 500 550 600 650 700 400 450 500 550 600 650 700 proposed by Sperling and WAVELENGTH (nm) WAVELENGTH (nm) Harwerth" or functions de- 25,000 22,000 19.000 16.000 25.000 termined for normal observ- 22.000 19,000 16,000 WAVENUMBER (cm-) WAVENUMBER (cm1) ers in a previous study using flicker methods.15

duces an equal reduction in sensitivity for all stimulus rous stimuli. To answer the question, spectral sensi- wavelengths (slope = 0.0002 ± 0.04, least squares cri- tivity functions were measured for the esotropic ob- terion). The differences between normal rivalry servers' deviating eyes during the suppression pro- suppression and strabismic suppression are particularly apparent in the short wavelength region of the spec- trum where rivalry suppression results in a dramatic reduction in spectral sensitivity.

1.2- Rivalry Suppression in Strabismic Observers 1.0-

0.8- When similar objects are presented to corresponding O CO O co points in the two eyes of observers with strabismus 0.6- of early onset, regions in the normally deviating eye 0.4- are almost constantly suppressed. When, however, 0.2- dissimilar objects, objects that induce binocular rivalry 0- in normal observers, are presented to the two eyes of 400 450 500 550 600 650 700 a strabismic observer with little or no amblyopia, the WAVELENGTH (nm) observer experiences an alternating suppression of 8 9 Fig. 4. Mean log ratio of the sensitivities obtained for three each eye. - Because the temporal characteristics and normal observers (filled symbols and solid line) and the five the phenomenal description of the alternating esotropic observers (open symbols) during monocular nonsuppres- suppression produced by rivalrous stimuli in normal sion conditions, and either the suppression phase of binocular and strabismic observers are similar, we wondered if rivalry (normal observers) or during strabismic suppression (esotropic our strabismic observers, like our normal observers, observers). The test probe dimensions were 20 msec and 0.85 deg. The sensitivity ratios obtained for one of the normal observers would demonstrate a wavelength-dependent loss of using similar procedures in a previous study5 are also included sensitivity when suppression was produced by rival- (filled symbols, dashed line).

Downloaded from iovs.arvojournals.org on 09/28/2021 86 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / January 1985 Vol. 26

Discussion 3.0- The similarity between the normal observers' in- crement thresholds obtained under monocular viewing conditions and those obtained during the dominance phase of rivalry indicates that the subjects' spectral sensitivity was essentially normal during rivalry dom- inance. However, in agreement with previous inves- tigations of increment thresholds during binocular rivalry,514 all the normal observers demonstrated some reduction in sensitivity during rivalry suppres- sion. More importantly, for test probe dimensions that normally biased detection toward the chromatic mechanisms, rivalry suppression resulted in a wave- length-dependent loss in sensitivity that was associated NORMAL OBSERVER with a change in the shape of the suppressed eye's spectral sensitivity function. If it is assumed that the 450 500 550 600 650 700 shape of the spectral sensitivity function indicates the WAVELENGTH (nm) type of psychophysical mechanism that dominates 25,000 22.000 19.000 16.000 detection, the alterations in the shape of the spectral WAVENUMBER (cm1) sensitivity function observed during binocular rivalry indicate that rivalry suppression in normal observers Fig. 5. Increment-threshold spectral sensitivity functions for the deviating eye of an esotropic observer for monocular viewing (open involves a selective reduction in the sensitivity of circles) and for the suppression phase of binocular rivalry (orthogonal chromatic mechanisms relative to luminance mech- gratings, filled circles). The stimulus subtended 0.85 deg and was anisms. As a consequence of the differential atten- presented in 20 msec flashes. The lower inset shows the log ratio uation of chromatic and luminance mechanisms, the of the sensitivity obtained during the monocular and binocular magnitude of the suppression measured with a test (suppression phase) viewing conditions. For comparison, the differ- ence function obtained for binocular rivalry for a normal observer probe will depend not only on stimulus wavelength is also included in the inset. but potentially on a variety of stimulus variables (eg, background luminance, stimulus size, and duration, etc.), which normally affect spectral sensitivity. duced by rivalrous stimuli (ie, orthogonally oriented In contrast to the mechanism-specific loss in sen- gratings presented to the left and right eyes were used sitivity demonstrated by normal observers during to produce suppression). Thus, in this experiment the binocular rivalry, the esotropic observers exhibited methods and stimuli were identical to those used to an approximately equal reduction in the sensitivity assess rivalry suppression in our normal observers. of both the chromatic and luminance mechanisms Figure 5 illustrates typical increment-threshold during strabismic suppression. Schor910 has shown spectral sensitivity functions determined for the de- previously that the optimal stimuli for inducing bin- viating eye of an esotropic observer under monocular ocular rivalry suppression and strabismic suppression viewing conditions (open circles) and during the are different. Our experiments quantitatively show suppression produced by the rivalrous stimuli (filled that the types of information suppressed during bin- circles). Although the strabismic observers reported ocular rivalry in normal observers and strabismic an alternating suppression of each eye when they suppression in esotropic observers are also different. viewed the orthogonally oriented patterns, the shapes Thus, it seems unlikely that normal binocular rivalry of the spectral sensitivity functions determined for and strabismic suppression are mediated by identical the suppressed, deviating eyes were identical to those mechanisms. obtained under monocular viewing conditions. The In agreement with previous investigations8'9 of lower inset in Figure 5 shows that the sensitivity of binocular rivalry in amblyopic subjects, the esotropes the deviating eye was reduced by the suppression reported an alternating suppression of each eye under initiated by the rivalrous stimuli, however, the reduc- rivalrous conditions. However, the alternating sup- tion in sensitivity was approximately constant for all pression process initiated in the esotropic observers stimulus wavelengths. Clearly the strabismic observers by the orthogonally oriented grating patterns did not failed to demonstrate the dramatic reduction in sen- result in a wavelength-specific loss of sensitivity. The sitivity for short wavelength stimuli characteristic of differences in the patterns of suppression demonstrated rivalry suppression in normal observers. in Figure 5 by the strabismic and normal observers

Downloaded from iovs.arvojournals.org on 09/28/2021 No. 1 BINOCULAR RIVALRY AND STRADISMIC SUPPRESSION / Smirh III er al. 87

under rivalrous viewing conditions indicate that the 2. Levelt W: On Binocular Rivalry. Institute for Perception, esotropic subjects did not experience normal binocular Rijksinstituut voor Oorlogsdocumentatie—Toegepast Natu- rivalry. Instead, the pattern of sensitivity loss exhibited urwetenschappelijk Onderzoek, Soesterberg, 1965, pp. 1-6. 3. Burian HM and von Noorden GK: Binocular Vision and by the esotropic subjects was identical for the parallel Ocular Motility. Theory and Management of Strabismus. St. and orthogonally oriented grating patterns. Therefore, Louis, CV Mosby, 1974, pp. 216-219. it appears that the esotropes manifested a form of 4. Dale RT: Fundamentals of Ocular Motility and Strabismus. strabismic suppression under rivalrous conditions. NY, Grune & Stratton, 1982, pp. 112-123. In summary, the results indicate that adults with 5. Smith EL, Levi DM, Harwerth RS, and White JM: Color vision is altered during the suppression phase of binocular constant esotropia of early onset do not demonstrate rivalry. Science 218:802, 1982. normal binocular rivalry and that strabismic suppres- 6. Davidson DW and Eskridge JB: Reliability of visual acuity sion is not simply a form of normal binocular rivalry measures of amblyopic eyes. Am J Optom Physiol Optics with different temporal characteristics. However, the 54:756, 1977. present study does not disprove the commonly held 7. von Noorden GK and Maumenee AE: Atlas of Strabismus. St. Louis, CV Mosby, 1967, pp. 84-96. view that binocular rivalry is an important phase in 8. Lawwill T and Meur G: Retinal rivalry in functional amblyopia. the development of strabismic suppression, ie, the Arch Ophthalmol 97:1100, 1979. 1 facultative suppression of Worth. 9. Schor CM: Visual stimuli for strabismic suppression. Perception 6:583, 1977. Key words: binocular rivalry, strabismus, suppression, spec- 10. Schor CM: Zero retinal image disparity: A stimulus for suppres- tral sensitivity, binocular vision, chromatic mechanisms, sion in small angle strabismus. Doc Ophthalmol 46:149, 1978. luminance mechanisms 11. King-Smith P: Visual detection analysed in terms of luminance and chromatic signals. Nature 255:69, 1975. 12. King-Smith P and Carden D: Luminance and opponent-color Acknowledgment contributions to visual detection and adaptation and to temporal and spatial integration. J Opt Soc Am 66:709, 1976. The authors thank H. Bedell for reviewing the original 13. Sperling HG and Harwerth RS: Red-green cone interactions manuscript. in increment-threshold spectral sensitivity of primates. Science 172:180, 1971. References 14. Wales T and Fox R: Increment detection thresholds during rivalry suppression. Percept Psychophys 8:90, 1970. 1. Duke-Elder S and Wybar K: System of Ophthalmology, Vol 15. Harwerth RS and Levi DM: A sensory mechanism for ambly- VI, Ocular Motility and Strabismus. London, Kimpton, 1973, opia. Part I: psychophysical studies. Am J Optom Physiol pp. 290-317. Optics 55:151, 1978.

Downloaded from iovs.arvojournals.org on 09/28/2021