THE EFFECT OF ‘INTERMI'ITENT PHOTIC STIMULATION ON VISUAL RESOLUTION THESIS FOR THE DEGREE OF PH. D. MICHIGAN STATE UNIVERSITY CHARLES M. BOURASSA I96 2 “masts This is to certify that the thesis entitled THE EFFECT OF INTERMITTENT PHOTIC STIMULATION ON VI S UAL R55 OLUT I ON presented by C mlRLES MARK BOURAS SA has been accepted towards fulfillment of the requirements for _Ehll._ degree in W W@' Major professor Date November 20, 1962 0-169 ABSTRACT THE EFFECT OF INTERMITTENT PHOTIC STIMULATION ON VISUAL RESOLUTION by Charles M. Bourassa Under the preper conditions sub—fusional intermittent photic stimulation produces a level of perceived bright- ness greater than continuous illumination of the same intensity level. This is called brightness enhancement (BE). To account for this and other related phenomena the alternation-of—response theory has been put forward. The theory maintains that with adequate target size and luminosity sub-fusional photic pulses tend to produce a large synchronous discharge of cortical neurons. It was hypothesized that a neuronal discharge of this type would also act to destroy the neural timing and interaction believed to be necessary for fine visual resolution. The hypothesis was tested by using two illuminated targets in an otherwise dark field. The targets, when overlapped provided a uniform visual surface. The targets were adjusted away from one another until the observers perceived a gap between them. By comparing the change in gap-size from continuous to intermittent illumination relative changes in visual resolution could Charles M. Bourassa be measured. Brightness matches were obtained by having the observers adjust a continuously illuminated target of the same size as the acuity targets to an intensity which would equal the brightness of the acuity targets. Intermittency was produced by two episcotisters which provided PCFs of 1/4 and 3/4. In general, five readings of perceived brightness were obtained at pulse rates producing fusion and at 20, 15, lO, and 5 pulses/second and also under steady illumination. Then five readings of visual resolution were obtained under the same conditions. Replications were made in many of the conditions. Medians of the five readings were used to present the data in graphical form. In general the findings may be summarized as follows. At intensities below about 1 c/ft2 the targets failed to produce BE. The perceived brightness produced by PCF 3/4 was always greater than the brightness produced by PCF 1/4. At these intensities visual resolution was positively related to perceived brightness in that visual resolution was better with PCF 3/4 than with PCF 1/4. At intensities between 1 and 10 e/rt2 PCF 3/4 provided for only small changes in perceived brightness. In this intensity range PCF 1/4 produced fairly large changes in brightness. With a target size of about 1016'26" BE occurred and visual resolution was correSpondingly worsened. With a target size of about 0030'33" the findings were similar, although a somewhat higher level of intensity was needed to produce Charles M. Bourassa BE. It was also found with this target size that when the intensity level was about 1500 c/ft2 no BE occurred. With a target subtending about 004'20" of visual angle it was found that increases in the intensity of continuous (uninterrupted) illumination adversely affected visual resolution. These small targets produced no BE but reductions in pulse rate adversely affected visual resolu- tion. In all cases in the present study BE was found to be maximal at 5 pulses-per-second. Whenever BE occurred the same stimulus conditions adversely affected visual resolu- tion. In some cases in which BE did not occur but where the intermittent stimulus conditions produce relatively large brightness changes, acuity is worsened. It was concluded that a high degree of synchronization of neural impulses, as reflected in changes in intermittent bright- ness, act to eliminate the fine gradations in neural timing or frequency which in ordinary circumstances provide a basis for some forms of perceptual discriminatimh . In some cases the changes in brightness produced by intermittency acted in the same way as did brightness changes produced by manipulating intensity. This is believed to occur when intermittent stimulation produces only a very small degree of cortical synchronization. Other aspects of the data were discussed and areas for further research were suggested. APPROVED:J.a<£....,5x.fz.v DATE: MLO.(“¥-Y o THE EFFECT OF INTERMITTENT PHOTIO STIMULATION 0N VISUAL RESOLUTION By Charles M3 Bourassa A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1962 G1 2.53500 OIIZIOE) ACKNOWLEDGMENT I deeply appreciate the continuous help, criticism, and patience of Dr. S. Howard Bartley during the formula- tion and execution of this study. I wish to thank the other members of my committee: Dr. W.D. Collings, Dr. Charles Hanley, and Dr. M. Ray Denny. Dr. Thomas Nelson and Mr. Richard Ball must receive thanks for their donation of time and effort, both as observers and critics. I alSO‘wish to gratefully acknowledge the many hours of time that my wife gave to this study in her role as data recorder and typist. ii TABLE OF CONTENTS Page ACKNOWLEDGEMENTS..................................... ii LIST OF TABLES....................................... iv LIST OF FIGURES.... ........ .......................... v INTRODUCTION......................................... 1 METHOD............................................... 28 Apparatus.......................................... 28 Observers.......................................... 30 Procedure.......................................... 30 RESULTS.............................................. 33 DISCUSSION........................................... 63 SUMMARY....................... ..... .................. 74 BIBLIOGRAPHY......... ......... ....................... 76 iii Table l. 3. 4. 5. LIST OF TABLES Page Resolving power of the eye under various conditionsooOOOOOOOOOOOOOOOOOOO0.00...0.0.0.. l (1) Visual angles subtended by the targets at two viewing distances..................... 30 Raw data acuity readings for observer RB....... 37 Raw data brightness matches for observer CB.... 37 Raw data from observer TN for acuity at various intensities of target illulninationOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 51 iv LIST.OF FIGURES Figure Page I. Perceived brightness of an intermittent photic source at various pulse rates......... 4 2. Relation between the Weber fraction and intensity of stimulation..................... 14 3. Visual acuity as a function of target intenSityooooooeo00000000000000.0000eooooooeo 15 4. Schematic diagram of the apparatus............. 28 5. Perceived brightness and visual resolution under steady and intermittent illumination for observers CB and RB at low intensity with target size of 1016'26"................. 39 6. Perceived brightness and visual resolution under steady and intermittent illumination for observers CB and RB at moderate intensity with a target size of 1°16'26"............... 4o 7. Perceived brightness and visual resolution under steady and intermittent illumination for observers CB and TN at a target intensity of .2 c/ft2 with a target size of 0°30'33"............................. 43 8. Perceived brightness and visual resolution under steady and intermittent illumination for observers CB and TN at a target intensity of about 1 c/ft2 and a target size of o°30'33"............................. 45 9. Perceived brightness and visual resolution under steady and intermittent illumination for observers CB and TN at a target intensity of about 5 c/ft2 and target size of O°30'33"............................. 46 10. Perceived brightness and visual resolution under steady and intermittent illumination for observers CB and TN at a target intensity of about 15 c/ft2 and a target size of o°3o'33"............................. 48 V Figure Page 11. Perceived brightness and visual resolution under steady and intermittent illumination for observers CB, TN, and RB at a target intensity of about 1500 c/ft2 and a target size of O°30'33"............................ 50 12. Perceived brightness and visual resolution under steady and intermittent illumination for observer CB at three target intensities; target size is 1046'50"; viewing distance is 100 cmOO000.000.0000000000000000000000000 52 13. Perceived brightness and visual resolution under steady and intermittent illumination for observer GB at three target intensities; target size is 004'20"...................... 54 14. Comparison of the perceived brightness of three target sizes at various levels of target intensity............................ 58 15. Combined brightness data for all observers under all conditionSOOOOOOIOOO0000.00.00.00. 6O 16. Idealized brightness curves for all ConditionSOOOOO00.0.0.0...O'COOOOOOOOOOOOOOO 6]- vi INTRODUCTION Preliminary statement.-The present study concerns the effect of sub-fusional intermittent photic stimulation on visual acuity. Briefly stated,the background of the study is this. To account for the increased brightness of a flickering source over that of a steady source of the same intensity, Bartley has put forward the alternation-of- response theory. This theory has not attracted wide notice, nor has it until recently been utilized in any way other than to account for brightness increases at sub-fusional flicker rates. Within the last few years however it has been demonstrated that sub-fusional flicker produces changes in saturation and hue of colored targets. Certain deductions from the alternation-of- response theory have been used to explain these phenomena. It is appropriate to ask whether this explanation will predict results in other situations. The measurement of visual acuity under conditions of sub-fusional photic stimulation provides an excellent testing ground since, as will be explained later, the predictions of the theory are entirely different than common sense expectation. It is this type of experimentation that will be reported here. In the present study such experimentation is carried out over a range of intensities and target 51268. 1 2 By this means it is hoped to present the data relating brightness, visual acuity, intensity, and flicker in a more comprehensive fashion than has been attempted hither- to, to deal more clearly with the possible mechanisms involved, and to relate the whole to the broader aSpects of brightness perception. Alternation-of—reSponse theory.-Before discussing the present study it is necessary to present information relevant to the alternation-of—reSponse theory. This discussion, when coupled with the description of acuity phenomena, will lay the foundations for the formulation of the hypotheses to be tested in the present work. It will also demonstrate that past work concerning sub- fusional stimulation and acuity has not only suffered from the use of inapprOpriate stimuli but that the experimenters were looking for effects exactly the opposite of those to be eXpected. It will also be pointed out that the use of acuity measurements provides a much more quantitative technique for the assessment of the perceptual effects of flicker than the technique used in the study of color changes under flicker. We have already explained that the alternation-of- reSponse theory was formulated to deal with brightness increases of sub-fusional intermittent photic stimulation (6,7,8,ll,12,13). The general sort of relationship reported in a number of studies is shown diagramatically in Figure l. The relative effectiveness of an intermit- 3 tent source is ascertained by matching a steady source to the intermittent one. When pulse rate is rapid enough the target is perceived as fused, i.e., steady. The point at which the intermittent target passes from one which is perceived as flickering to one which is perceived as steady, or vice versa, is known as the critical flicker frequency (OFF). The brightness of the inter- mittent target at OFF is determined by the percentage of the stimulus cycle (a cycle being the time from onset of one photic pulse to the onset of the next) occupied by the pulse. The percentage of the cycle length occupied by the pulse is represented by the pulse-to-cycle fraction (PCF). A PCF of one-half means that half the stimulus cycle is illuminated and half not illuminated. In Figure l we have assumed that the PCF is one-half. The brightness at OFF, i.e., Talbot level brightness, is one-half of the brightness of a steady target of the same intensity. If the PCF were one-quarter, the Talbot level would be one-quarter of the steady. In other words, the brightness at fusion is equivalent to the brightness that would be produced by Spreading the photic pulse evenly over the entire cycle length (7,44,51). As shown in Figure 1, when pulse rate is reduced brightness begins to rise above the Talbot level and eventually exceeds the brightness of the steady target. The area in which brightness of the intermittent target is above Talbot's level, but below the level of the steady target, is referred to as the intermediate range (IR) of brightness (l6). .Brightness enhancement (BE) occurs when the intermittent target exceeds the brightness of the steady target. The pulse rate which produces the maximum enhancement, subject to the qualifications to be discussed shortly, is around eight to twelve per second (6). 2 t3 m 11. ll. 1.: m Brl htness of g, I :2 <1 E Talbot level CFL_ o 1 1 . 1L . 1 n 1 1 1 1 m 4 1 l . 1 1 ) 4 8 I2 l6 202428 32 36 4O PULSES/SEC Fig. l.--The general relation between pulse rate and perceived brightness when the source of illumination is interupted by an episcotister with a PCF of .5. See text for a complete account. It should be pointed out that while lower PCFs have lower Talbot levels of brightness they produce correspond- ingly greater BE (7,16). To account for the phenomena of Figure l the alter- nation-of-reSponse theory was put forward by Bartley (6,8, 11,12). We may now briefly summarize the salient features of this theory. The alternation-of—reSponse theory maintains that there are separate and relatively indepen- dent parallel channels from retina to cortex. Each of these channels has its own latency, thresholds, and recovery time. When stimulation is maintained for some time the channels will recover and fire again in a 'random' fashion. When stimulation is intermittent each pulse of radiation will activate a certain number of channels. If the photic pulse is a relatively long one then, after an initial discharge of the retinal elements, some of the channels will recover and fire again as previously described. If the photic pulse is shortened there will occur an interval in which the photic pulse will be long enough to activate a large number of receptors but the pulse will go 'off' before the channels recover. Further, if the repetition of the pulse is at the prOper rate the next pulse will fire a number of recovered channels. Not all channels may have recovered by the time the second pulse is delivered but by the time the third pulse occurs some of the lagged channels plus perhaps a few of the channels which recover rapidly from the second pulse will fire. Channels will thus rotate activity so that while no one pulse fires all possible channels each pulse will fire the maximum number of recovering channels thereby allowing for maximal utilization of the retino-cortical pathway. Furthermore, neurophysiological investigation has disclosed that certain rates of delivery of nervous impulses to the cortical visual system result in a maximal cortical response (4,7,8,36,38). Within the 6 framework of the alternation-of reaponse theory this latter point is of prime importance since, No matter what the discharge rate in the retina happens to be, the imput has to be favorably received by the cortex before sensory effects will be maximal and be expressions of brightness enhancement (12). It must be pointed out that not all conditions of stimulation produce the results shown in Figure 1. From the alternation-of-reSponse theory it is possible to predict that low target intensities and/or small target sizes may reduce or eliminate BE. This follows from the fact that small targets or low intensities are not able to fire enough channels to result in a large cortical response. This is, of course, in accord with information concerning the relation of visual acuity, brightness discrimination, and CFF to changes in target size and/or intensity(7,44,51,b7). It has also been shown that with some combinations of target size and intensity, brightness of the intermittent target will continue to increase below ten pulses per second (p/s),i.e., intermittent brightness reaches a maximum at pulse rates below 10 p/s (10,21,22). Unfor- tunately such studies have not been systematic enough to disclose exactly the crucial parameters influencing the results. There are other factors which will also influence the BE function. The effect of FOP has already been briefly described. According to the alternation-of—response theory the advantage of low PCFs lies in the assumption 7 that smaller PCFs tend to produce a more uniformly maximal reSponse, whereas longer PCFs allow enough time for some of the initially stimulated channels to go out of action before the pulse is ended. Another major factor thought to influence BE is the amount of stray light present on the retina. If tne stray illumination is supra-liminal it will stimulate some of the neural channels which would otherwise be utilized by the photic pulse (17). Yet another factor considered to influence BE is the rate of onset and rate of decay of the photic pulse. There has been no systematic eXploration of this variable but since wave form can vary from square to an elongated sine we can eXpect some influence on BE. There is no way to hypothesize the exact nature of this influence without eXperimental data, although we might eXpect square pulses to produce a maximal neural response (37). We must also mention a concept which can be deduced from the alternation-of—reSponse theory and which has received substantiation from both physiological and psychological experimentation. This concept is that of the cortical reorganization period and it states that before rythmic stimulation reaches full effectiveness a certain amount of time must elapse in order to allow the neurons to group reaponses (15). That is, there must be a period of time for the various channels to fall into line and organize the reSponses. Functionally, this means 8 that intermittent stimuli must be observed for some small but appreciable time before the rythmic nature of the impingements produces brightness enhancement. Thus we must not eXpect extremely short exposures of the inter- mittent target to result in BE. This brief account of the data concerning BE and of the mechanisms proposed to account for BE may be suppli- mented by reference to various papers already cited in which a more complete account is given. Expansion of alternation-of-reSponse theory.-For many years the alternation-of—reSponse theory has been limited to an account of sub-fusional brightness changes. Recently however Nelson and Bartley have utilized this theory to help eXplain the wash-out or desaturation of various hues viewed at sub-fusional intermittency rates. For details the reader may consult the original publications (14,53, 55). We may summarize the matter here by saying that it is hypothesized that the rythmic neural firing underlaying BE destroys the timing of the neural impulses which determine the perception of color. Perhaps because of the strong emphasis placed on cortical activity by the alternation-of-reSponse theory, the authors have emphasized cortical timing as the major factor in wash-out. Whatever the locus of the mechanisms, the alternation- of-response theory has been tentatively extended to deal with phenomena other than brightness inoreases. A major deficiency of the wash-out research lies in the fact that 9 for various good reasons, the investigators were unable to use a matching target. Thus the reports of desaturation and brightness changes are largely qualitative. While there is no doubt that desaturation phenomena occurs there may be some doubt that the conditions used produce BE. It is, moreover, clear that the wash-out takes place over a much wider range of pulse rates than does BE. There is no question that use of a matching source would allow a much stronger case to be made. It is to be eXpected however that desaturation is likely to occur before BE. That is, brightness perception seems to be a primitive reSponse to stimulation and it is likely that relatively larger stimulus changes will have to occur before brightness changes take place whereas Spectral discrimination will be affected by smaller changes in the conditions of stimulation. Thus wash-out may be a more sensitive indicator of alterations in neural timing than brightness. Visual_resolution.-Before extending the alternation-of- reSponse theory to problems of visual resolution it is necessary to briefly discuss visual acuity. Acuity is expressed as the reciprocal of the minimum visual angle, in minutes, which the observer is able to resolve under a given set of conditions. Thus as the numbers representing acuity get larger, visual acuity improves. However, it is usually the minimum resolvable visual angle which is measured, and as this angle increases acuity decreases. 10 The difference between acuity and minimum angle of resolu- tion (MAR) must be understood because both terms are used here; since, as Ludvigh and Weymouth point out, the . minimum angle of resolution is of primary importance (69), most of the experimental data presented later are plotted in terms of MAR. We may begin this examination by considering the perception of a point source. Although in reality no source can be a point, since a point has no area, we may for the present accept the common definition of a point source as one whose distance is very great in relation to its size (28,48,64,66). When a point source is barely liminal the observer perceives a dimensionless point. As intensity is raised the perceived point gains radial offshoots, at first short and then longer (58). The radial offshoots are due to inhomogeneities in the refracting media of the eye. Such inhomogeneities are often classed as aberrations (Spherical aberrations, astigmatism etc.) (28,42,48,64,66). It should be noted that such phenomena are aberrations only in terms of an already articulated theory of optics and are not necessar- ily aberrations in terms of visual perception (58). The reader may also consult Walls who points out that many of the so-called aberrations do not hinder, but may aid perception (68). If the intensity of the point source remains constant and the surrounds are gradually made luminous, perception ll of the point is at first unaffected, but it will eventually lose the weakest rays and the longer ones will shorten. This occurs, in part, because the increased illumination results in a decrease in pupil size which will reduce some of the irregularities of the ocular media through which the radiation must pass (29,42,51,58,64,68). Reducing the aperature will also act, initially at any rate, to decrease the retinal illumination hence causing some of the radiation to become sub-liminal. If we resume increasing the intensity of the point source the reSponding retinal area will be reduced because the pupil contracts and because retinal sensitivity will decrease. On the other hand, the increasing amounts of entOpic stray radiation act to increase the affected area. Over an intermediate range these two factors balance (22, 42,53,58), but eventually pupil size and sensitivity reach minimum values and further intensity increases result in dazzle or glare which reduces visual efficiency (53). This effect will occur when the target or the surround luminosity is too high and also with either point or extended sources (53). In order to further simplify this discussion let us now consider the effect of small artificial pupils. A diameter of less than one mm will eliminate the irregular- ities of most eyes (58). With a small artifical pupil the energy is distributed in a diffraction pattern with a central disc of maximum brightness and rings of decreasing 12 intensity (53,58,64). Following Ronchi the term 'centric' will be used to refer to this diffraction pattern. The intensity of the rings of the centric are considerably less than the intensity of the central disc. Thus, while at high intensities five or six rings may be visible, at lower intensities only the central disc is perceived (53, 58). The diameter of the central disc of the centric is inversely related to the diameter of the aperture (53,58, 64). Consideration of this relation might lead to the prediction that, since the Optical image is 'sharpest' when the centric is a minimum, enlarging the aperture should produce better vision. While this is true in certain cases, e.g., the mirrors of reflecting telescopes, we have already pointed out that increasing pupil diameter uncovers more irregularities in the refracting system of the eye. Cobb was able to demonstrate that over a midrange of pupil sizes the two effects cancelled one another, i.e., acuity was unchanged over a range of pupil sizes from about 2.5mm to 4mm (22). If, by using two artifical pupils, centrlcs are placed so as to coincide on the retina the only effect is to increase the brightness of the centric (58). If the centrics are placed so as to overlap slightly then bright- ness will increase and the observer is also likely to report that the centric is no longer round but oval (58). If we continue to add centrics, at a distance from one 13 another small enough so that no gap is perceived, both brightness and 'ovalness' increase. After a certain number of additions is exceeded brightness remains constant, i.e., the addition of a single centric does not contribute to the original flux, but the 'ovalness‘ becomes perceived as a line (58). Thus when reduction of a source produces only a brightness change it is, for that observer, a point source. We can also define linear, and by extention, extended sources (58). Turning more directly to acuity we note that visual resolution means perception of two point sources with a darker area between them. The affected retinal zone is determined partly by intensity, but also by diameter of the artifical pupil, wavelength of the radiation, retinal threshold, etc.. In any particular case these factors can be held constant, but then we must know the sensitivity of the eye to differences in intensity. Figure 2, redrawn from Ronchi, shows the reciprocal of Webers fraction plotted against intensity (58). The dependence of acuity on intensity can be represented by a curve which goes down at either end. It drops at high intensities because this leads to a widening of the centrics disc and because of the increase in the Weber fraction. The curve should drop at low intensities since Webers fraction increases and also because the stimuli become subliminal. The same considerations by and large apply to extended surfaces except that increased intensity will not lead to L 14 the same Spread of the 'image'. At high intensities a linear source is seen much wider than it 'really' is (30,59). An extended source will obviously not show the same percentage increase in apparent width (30,58). .1. Ill «40- 30L 20‘ I L 1 1 '4 '2 O 2 4 Loql Fig. 2.--The relation between the Weber fraction and intensity of stimulation. From Ronchi (58). See text for details. Turning now to actual data, Figure 3 shows results of several experiments relating acuity to intensity. The curve which levels off at high intensities is Konig's. The leveling off is presumably because no surround illum- ination was used. The other curves, with the exception of the Wilcox, and the Fry & Cobb curves, were obtained with the surround illuminated at the same level, except at the highest intensities, as the targets. Wilcox' curve was obtained using very narrow bars on a dark background (70). The visual angle subtended by the targets was about 142". The relationship here is similar to that to be eXpected of 15 2‘6 Eguéhi 2.4 / ’ /", Lyt goe 2.2» /,’ 2.0 / Fry _, 8. LB A Cobb(Wi // Koni l.6 // ///Cob 8. Moss _. ‘ 1 // / l.0 /// 77% / // 2: / y 0.4 / / Wilcox / 0.2 l.4 l.2 Acuity K. \ i I L :04 l0" Io-3 I0"2 ID" I l0 no2 lo3 10‘ lo5 Luminosity of Background (Lumens/fizi 0.0 Fig. 3.--Visual acuity as a function of target luminosity. The curves of Eguchi, Lythgoe, Konig Cobb & Moss, and Konig & Heymann are taken from Moon (51). The curves of Wilcox, and Fry & Cobb are taken from Bartley (7). See text for details. 16 a linear source. The curves of Fry & Cobb labelled N and W directly show the effect of target size when using illuminated targets in a dark surround(7). The N curve was obtained with targets subtending a visual angle of 168" while the w curve was obtained with target subtending 1000" of visual arc. Thus the data bear out the principles discussed earlier. The physiological mechanisms of visual acuity are largely unknown. Several theories have been put forward from time to time and for a discussion of the most prominent of these the reader is referred to papers by Hecht, Hartridge, and Shlaer (42,44,62,63) or to the other sources listed in this section. Recent neurOphysiological investigations of the retina and Optic nerves demonstrate that vision is a much more complex affair than the earlier theorists realized (2,3,7, 29,35,36,39,4o,41,47,52,56). The retina, like other sensory end organs, is organized in complex sets of overlapping receptive fields. Presumably there is interaction not only within but also between receptive fields (3,6,35,39,40,4l). Radiation on the retina does not start a simple neural message to the cortex. Moreover, the retina and the optic cortex are Spontaneously active. Any neural impulses originating from external stimuli must be considered in relation to the ongoing activity. It must also be borne in mind that while the foveal area is limited to a small portion of the retina, foveal representation is found to 17 take up the largest part of the visual cortex (1,35,59,68). Without going into detail current thought indicates that the interaction of receptive fields acts to sharpen gradients of stimulation (35). The role of Spontaneous activity, by analogy with its role in other receptors, lies in its ability to indicate by deviations in either direction a more complete picture of retinal stimulation (35). That is, either an increase or a decrease in spontaneous activity is significant for perception, So far we have been concerned with retinal mechanisms. We must now ask about cortical influences. Neurophysiologi- cal investigations of cortical 'contour processes' have yet to be carried out but there is some evidence that the processes just described in the retina are also active in the cortex (52,56). At least in man the cortex is essential for visual discrimination (59) and it seems to have been adequately demonstrated that form perception is at least partially dependent on cortical mechanisms. Evidence for this has been presented by Hebb and others (43,59). Many vertebrates, other than man, are able to discriminate light from darkness, and differences in total retinal flux (not brightness) when the cortex is eliminated (7,59). There is some justification for believing that difference in total flux is the most primitive sort of visual reSponse (21,66). As we have already indicated, form perception seems to involve a more elegant mechanism. Presumably the perception of a 18 straight line falls at some intermediate point in terms of the complexity of the mechanisms involved. This assumption follows from the reports that Senden's subjects were able to distinguish an object from a ground but not able to deal with the whole object. Therefore they must have perceived some contours but were not able to follow them (43).. Table 1, taken from Moon (51), gives some average acuity Values for various conditions. As might be eXpected the values from the various conditions differ widely. One must remember however that the conditions are not directly comparable. That is, target size, target illumination etc. cannot be equated in the various conditions so that in reality the acuity values would be eXpected to differ. TABLE l.--Resolving power of the eye under various conditions APPROXIMATE RESOLVING POWER OF THE EYE Object Minimum visual angle Distance Visual Minutes Radians on retina acuity Two stars 1.0 2.9 X 10"4 4.3microns 1 Two black bars 0.5 1.45 2.1 2 on white ground Vernier 0.15 0.43 0.65 6.67 Spider web or black wire on 0.13 0.38 0.57 7.69 self-luminous background 19 We might ask however whether the differences in acuity values obtained by the various techniques are to be attributed solely to differences in energistic impingements. The contention here is that some of the differences arise because of a difference in physiological mechanisms employed in the discrimination. Such Speculation is of course idle unless some method of testing it is found. It may be that sub-fusional intermittent stimulation will differently affect these acuity meaSures. Specifically, we expect the more elaborate discriminations to be more seriously affected by synchronous nervous activity than the more primitive responses (see section on extension of alternationéof response theory). A test of this hypothesis is outside the sc0pe of the present work but indicates one interesting area for further investigation. Effect of intermittent stimulation.-Predicting the effect of intermittent stimulation on visual acuity is somewhat difficult due to the lack of knowledge concerning the mechanisms of acuity. It is also difficult to isolate some portion of the visual nervous system as the locus for a.phaxmwnon for disruption of any one part of the system is likely to affect other parts. As Hylkema says, For a good functioning of the organ of vision it is essential that all the links in the chain be sound, and this holds good for all the qualities of sight. It is generally admitted that visual acuity depends upon the eye (distinguished in refractraction system, photosensitive cells and retinal nervous tissue reSpectively), as well as on the nerves connecting it with the cortex cerebri and on the cortex itself.... But which organ plays the chief part cannot Simply be 20 stated without anything further (46). It may be possible however to point out by analogy the probable effect of intermittent stimulation if we examine some other receptor reSponse system about which more is known. One such system involves the muscle tendon afferents. The visual receptors, i.e., rods and cones, are of course much different in structure than the muscle Spindles and tendon organs, and the adequate stimuli for the two sets of receptors are very unlike. None the less the basic organization seems very similar (33,59). The afferents of muscle Spindles and tendon organs converge on ganglion cells and are arranged in overlapping receptive fields. They are Spontaneously active and taken together give responses similar to the 'on' and 'off' responses of the retinal ganglion cell. The muscle spindles and tendon organs 'adapt' to various states of tension just as the eye adapts to various levels of illumination. We might note however that Spindle adaption to tension comes about due to efferent impulses changing the 'loading' on the receptors. Presumably the eye is little effected by efferents, although this presumption may well have to be modified. In the normal course of events the tendon organs and muscle spindles act to aid in providing smooth, integrated muscular responses. When however a sudden strong stimulus is applied (a blow to the patellar tendon) the result is a more or less violent response. This comes about because 21 of a large scale synchronous discharge of the receptors, which eliminates the fine reSponse gradients occurring under ordinary circumstances (59). It seems likely that applying sudden strong stimuli to the visual system must result in a similar state of affairs. That is strong repetitive stimuli prevent the fine gradiations of neural timing and interaction necessary for a well ordered, highly differentiated reSponse. Such an analogy might be extended further but it seems unnecessary to do so at this time. It is worth repeating however, that apparently in the visual system 'primitive' reSponses, such as perception of straight lines, movement etc. might be less affected by intermittency than more 'elegant' reSponses such as hue and saturation. Previous work.-Before stating the hypothesis in an exact manner we may inspect previous work relating acuity to flicker. Aside from the compelling reasons provided by the alternation-of-response theory and the work on washout, impetus for studying the relation between acuity and flicker has come from Orozier's work on the influence of the avian pectin on movement acuity (26). In the course of checking some of his findings in humans, Orozier reported that the intensity needed to perceive a pattern remained the same deSpite variations in FOP, although these manipulations of FOP caused large brightness changes (27). Crozier further pointed out that flicker studies might allow the disentan- 22 glement of brightness and intensity; the importance of studying the relation between form perception and flicker was emphasized (26,27). Following up the suggestion of Orozier, Senders (60) and later Nachmais (52) reported essentially similar findings, i.e., that at low intensities of illumination low PCFS provided for resolution of a grating at lower intensities than high PCFS. Nachmais demonstrated that brightness followed the Talbot level regardless of pulse rate. The two studies just cited were concerned with BE in only an incidental fashion. The target intensities were too low for enhancement to occur. However we might note two assumptions made about BE. Both investigators believed that if BE did occur it would be maximal at about 10 p/s which as we have seen is not always true. In any case Nachmais's brightness matches dispose of the possible cncurrence of BE. It may be noted in passing that Nachmais is the first to report that brightness matches at low intensities follow Talbot's law regardless of pulse rate although it must be added that Bartley, using a disc- annulus arrangement, has shown that at low intensities of intermittent stimulation, brightness may drop below the Talbot level (9). Aside from the aSiumption that BE will be maximal at about 10 p/s even at low intensities, both authors also believed that increasing brightness by means of intermittent stimulation would act in the same fashion as a brightness 23 increase produced by raising intensity; i.e., to improve perceptual discrimination. This belief is based on the assumption that the mechanism of brightness perception, no matter how activated, will always produce the same end results. We have seen earlier that this is not necessarily the case. Apparently only one study has been reported in which the aim was to directly assess the effect of BE on form perception. In this study Gerthewohl and Taylor assumed that if BE were Operating the increased brightness should improve their subjects performance (31). To test this hypothesis they utilized a chart in which the print started out in the first line as black on white, and then, through succeeding lines the contrast decreased; i.e., the back- ground gradually became darker. The number of lines read was used as a measure of the effect of various conditions. Using four POFS varying from .4 to about .96, with the source at about 15.4 foot-candles and pulse rates of nine and fifteen pulses-per-second, they found that there was virtually no difference in number of lines read under the various flicker conditions. Steady illumination always produced the best results. The number of lines read at the Talbot level of brightness however was only slightly greater than the number read under the flicker conditions. Gerthewohl and Taylor concluded that the Bartley effect was not Operating Since if it were the pulse rates near 9 per sec should have produced increased 24 brightness, and therefore, according to their line of reasoning, also improved the subjects performance. No brightness matches were actually obtained. This was a badly mismanaged experiment and it is perfectly clear that the authors did not at all understand the nature of the phenomaxn they set out to investigate. We have already pointed out that the initial hypothesis is likely to be wrong. Also the level of illumination was relatively low to begin with and, since the reflectance of the chart decreased line by line, target illumination became quite feeble. Hence if BE did occur, which because of the low illumination is not likely, it would make itself felt at some pulse rate below 9 per sec. Moreover the fact that Gerthewohl and Taylor did not obtain bright- ness matches is almost inexcusable in an experiment where perceived brightness is supposed to be the major factor under investigation. Senders examined the data of Gerthewohl and Taylor and claimed that their findings supported her own in that low PCFS did not reduce acuity more than the higher PCFS (61). Present study.-With this background information in hand we may proceed to a discussion of the present study. It is well known that an increase in target intensity tends to improve most visual functions. Usually target intensity is increased by increasing the amount of luminous flux reaching the eye from the target. This usually results in 25 the target being perceived as brighter. AS we have seen some experimenters have assumed if the target is made brighter by means of BE that this should also improve visual discriminations. From the material discussed in the previous sections of this paper we know that this assumption is erroneous. In fact, if BE occurs, visual discrimination should become worse. Thus there exists here a clear cut difference between theoretical and common sense understandings of BE which may be put to test. In the present study, the effect of intermittent photic stimulation on visual acuity is to be measured. Such measurements will prove to be significant in a number of ways, one of which was mentioned in the preceeding paragraph. It will be possible by the Selection of proper stimulating conditions to systematically vary the factors of target size and intensity, which as indicated earlier, are major factors in BE and also, since brightness matches are to be directly obtained, it will be possible to state with certainty, and in a quantitative fashion, the relation Of BE to acuity. Moreover, it will be possible to separate with some certainty effects due to impingements per second from the effects associated with BE, thus allowing a more positive statement of the relationship of BE and acuity than it has been possible to make concerning BE and washout. In order to be able to obtain meaningful brightness matches the type of target must be chosen with some care. In this case it was decided to use two illuminated targets 26 in a dark field. The targets could be adjusted toward or away from each other. The size of the interSpace which the observer found necessary in order to perceive a separation between the targets provided a direct measure of visual resolution. This method was employed by Wilcox who found that with extremely narrow targets increasing brightness caused an increase and then a decrease in acuity (70). This effect does not occur with larger targets (7). With this technique direct measures of acuity under a wide intensity range may be obtained. Brightness matches can be obtained by use of a second target equal in size to the combined area of the acuity targets. Although the hypotheses to be tested have been implicit throughout we may now state them in a formal manner. The following predictions are made: 1. Small target sizes combined with low intensities will not produce BE at any pulse rate. 2. Combinations of moderate target size and intensity will produce maximal BE at a pulse rate lower than 10 per second. 3. Large targets combined with high intensities will produce maximal BE at 10 pulses per second. Concerning the effect on visual acuity of sub-fusional flicker, we have already mentioned that the type of acuity measure used is very close to involving 'pure’ visual resolution and is very much akin to, in Hebb's terminology, a primitive figure ground perception. Thus it is likely to be much less effected by neural timing than hue and we therefore make the following predictions. 27 1. When BE does not occur there will be little difference between acuity measured at various pulse rates. 2. When BE does occur acuity will be worst at the same pulse rates that result in BE. METHOD APPARATUS The apparatus is shown diagramatically in Figure 4. E Lzl—4 F2 T2 TI Fl L3 LI 8:101 v "8 Fig, 4.--Schematic diagram of the apparatus. 3 are light sources. Ll, Lg, L3 & L4 are lens. represent filter holders. T1, T9, & T3 are fir mirrors used as targets. The episcotister is observer is seated at O. The sources (31, 32) are 200 tht frosted lamps contained in lamp houses. The lens L1 and L4 are used to collimate the radiation from the sources. Lens L2 and L3 could be used t bring the beam of radirtion to a focus. L3 is used to focus the beam at the plane of the episcotist- 29 er E. The targets (T1, T2, and T3) are first surface mirrors of various sizes. The intensity of the illumination falling in the targets can be controlled by Variac and by inserting Opal glass filters at F1 and F2. The entire apparatus, excepting T1, is rigidly mounted. T1 is fixed to a lathe micrometer head in such a way that movements of the micrometer slide T1 to the right or left.. Measure- ments of movement could be made easily to the nearest .02 mm. Target sizes.-Three target sizes are used. The smallest targets are each 4 mm wide. The medium size targets are each 22 mm wide and the largest targets are each 5.5 cm wide. In order to provide uniform illumination of the large targets it was necessary to modify the apparatus in the following way. Project-c-chart projectors are used to illuminate plane mirors which are placed at the 31 and 32 positions in Figure 4. The radiation is then reflected to the targets. The lens are not used in this case. The visual angles subtended by the targets, each target alone and by both targets combined, i.e., T1 and T3, at viewing distances of lOO a> - 1 -4I.. : : -: l l I Al I . 1 B] a I l I l ._ C §:£3£)h- - 9 (al.5- - (D u] E I.O~ - J: 9 «0.5- - — - CD ()1) I I I I I I I _I S 5 l0l5 20 S 5 l0l520 PULSE RATE Fig. 8.--The c and d portions of the figure show the brightness matches given by the indicated observers, under conditions of steady illumination (3), and under intermittent illumination at 5, 10, 15, and 20 pulses/sec. The dots represent data using a 1/4 PCF, the crosses 3/4 PCF. The a and b portions show, for each observer, the change in minimum angle of resolution (MAR) necessary in order to perceive a gap under the intermittent stimulus conditions as compared to MAR under steady illumination. MAR is in seconds of visual angle. 46 I I I I “133'2 F - >'CJ P - 258- - J - - (r “J _ <4- “2 C) a I I A»;- EE I T r I