The Integrative Action of the Nervous System

It appears to follow from such considerations that when we find electrical excitation of certain spots of the cerebral hemi- sphere regularly evoke certain movements, e.g. of a limb, the probability is that we have there a nodal point where various harmoniously acting neural arcs are tied together and can be there reached and driven as a unit — though a highly syn- thesized one — and produce the effect which is the common resultant of them all. The receptive points and organs which under stimulation initiate reflex movements also initiate, in the intact animal with unmutilated brain, sensations. As each reflex has a reflex action attributive to it, so it has potentially at least a sensual reaction. These sensual reactions, like the motor reflexes, are of various grades of complexity. The simple perceptual image of an object is usually a resultant as regards external stimulation of stimuli applied jointly to several sense-organs. It has direct sensual factors traceable from various sources. The cigar taken from its box may be simultaneously sensed through eye and hand and nose and ear. The object experientially regarded as a single object excites a neural reaction that has its starting points in many spatially and qualitatively distinct receptive points, each point the commencement of separate nervous arcs. In the psychical result of the reactions thus set going there is amplification and modification by conditions memorial, affective, judicial, conative, etc., obtaining in the mind and not due immediately to the stim- ulus. The neural process resulting from the nervous impulses

litiated by the retinal, olfactory, cutaneous, and muscular recep- >rs is therefore internally modified in the nervous system by processes and states already existent there or evoked there by jelf as a reverberation of its action. Can we at all compare rith the simultaneous co-ordination of the nervous factors in a lotor reflex the synthesis of the nervous elements whose com- Mnation underlies a simple sense-perception ? We may somewhat reduce the complexity of the sense- ►ercept by limiting its paths of initiation to those of a single inse, namely the visual, excluding the object, e.g, the seen igar, from directly stimulating other sense channels, tactual, olfactory, auditory, muscular, etc. The cigar may be offered mly to the eye. Then we have left as regards the external Emulation merely the fusion of the right-eye and the left-eye lages. This fusion is so complete that we cannot by intro- jection discriminate in the visual image the right-eye image from the left-eye image. Moreover, this fusion is so elemental that introspection cannot detect in it any effort of memory, judgment, or reason. It appears innate — a datum ready pro- vided even at dawn of individual human consciousness. We :an further strip the problem of some of its complexity by sub- 'stituting for the three-dimensional object, e. g. the cigar, with its perspective shading and its patches of colour and its char- acteristic associations, etc., a simple and relatively meaningless discoid patch of moderate, even, and uncoloured brightness, small enough to lie wholly on the central area of each retina. We can then test to what degree the visual singleness of the observed surface, sensed through right eye and left eye together, is due to direct confluence of the sensory paths excited by the right-eye and left-eye images respectively. We can attempt this in the following way.^^^

A double sheet of thick milk-glass is observed by transmitted light given by a lamp. This lamp is set in the axis of a rotating cylinder (Fig. 'j6). In the side of the cylinder are three horizontal rows of rectangular windows, tier above tier. The lamp, though fixed in the axis of rotation of this revolving cylindrical screen, is entirely free from all attachment to it. The milk-glass plate Figure 76. — Rotating Lantern. I. Elevation seen from front. II. Horizontal plan, through level of A-A of I. Supports seen in perspective. The eyeballs, pupil screens, and convergant visual axes are indicated belonging to II, but carried through I. The plan of the lantern is given one fourth actual size. Description in text.

is fixed between the lamp and the inner face of the tiers of windows, close to the latter. Outside the moving cylindrical screen is a fixed semi- cylindrical screen concentric with the revolving one, and just wide enough to allow the inner revolving one to turn within it freely (Fig. 76). In the fixed cylindrical screen four circular holes are arranged so that two are centred on the same horizon- tal line, and of the other two one is centred just so far above the left-hand hole of the just mentioned pair as the other is below the right-hand member of the pair. The horizontal distance between the centres of the right and left hand holes is 9 mm. The diameter of each hole is 8 mm. The vertical distance

between the centres of the holes is exactly the same as that be- tween the centres of tiers of the revolving cylindrical screen, namely 1 1 mm. These four circular holes in the outer fixed cylindrical screen are, in the experiments, viewed from a distance such that when the line of visual direction of the right eye passes through the centre of the right hole it meets (Fig. 76) at the axis of the cylindrical lantern the line of visual direction of the left eye, which latter Hne passes through the centre of the left-hand hole.

This being so, the images of the lower left-hand hole and of the upper right-hand hole fuse visually to singleness. They then appear as the middle one of three arranged vertically one above the other. A black vertical thin screen set at right-angles to the plane of the forehead is introduced (Fig. ']']) between the eyes and the holes so as to screen from the left eye all view of the right-hand holes, and from the right eye all view of the left-hand holes.

The revolving screen is driven by an electromotor. The speed of revolution of this motor is controlled by a coarse ad- justment and by a fine adjustment. The speed of rotation of the cylindrical screen is recorded by marking the completion of each revolution of its spindle by an electromagnetic signal writing on a travelling blackened surface (Fig. 'j']). On the same surface the time is recorded by a writing-clock marking fifths of seconds. The inner revolving screen by its revolution opens and shuts alternately for equal periods the circular holes in the fixed outer screen. The inner screen with its three tiers of windows is made in three pieces, each containing one tier of the windows. The piece containing the middle tier of openings is jointed in such a way that its openings can be set at any desired interval with the openings of the lowest tier. The highest tier is similarly jointed to the middle tier. In this way it can be arranged that the uppermost circular hole is open when the lower ones are closed, or is shut when the lower are closed, or is opened to any desired degree either before or after the lower; further, by removing the top gallery of the rotating screen it can be left permanently open. A similar relationship is also allowed between the middle holes and the lower.

By wearing weak prisms with their base-apex lines vertical the images of the right-hand and left-hand holes can be brought to the same horizontal levels. The observer can then immedi- ately fuse the four images to two by convergence. A horizontal fine thread halving each of the two middle holes, and similar but vertical threads halving the other two holes, serve to certify binocular vision to the observer. When the four holes are all allowed to act thus under appropriate convergent binocular gaze they are seen by the observer as two evenly lighted discs, one vertically above the other, and each cut into quadrants by a delicate black cross. By separately adjustable shutters any one, or any vertically edged fraction of one, of the discs can be sep- rately screened out of vision.

The object of the above arrangement is to attain the fol- lowing conditions. Images accurately similar are received by retinal areas fully visually conjugate. The areas are not only of the so-called " geometrical identity," but are at the time of the observation in full binocular co-operation, owing to the con- current convergence and accommodation. Extinction and illu- mination of the images occur pari passu in the two eyes, i. e. with like speed and in like direction. It can be synchronous or of any time-sequence desired. That the speed shall be similar for the two is insured by all the shutters being on the same spindle.

Each disc-shaped image will have on the retina a diameter of about 570/Lt. That is, when foveal vision is directed upon it, the image will occupy a practically rod-free area containing about 2,800 cones. The direction of translation being the same for all the shutters the bright images on the two retinae are, if the shutters are set for simultaneous right and left images, com- menced on " identical " points of the two retinae, established progressively along *' identical " points, and finally extinguished in like manner progressively along ** identical " points. Or, con- versely, if the shutters are set for accurately alternate right and left images the screening off begins in one eye at a spot and moment identical with those at which the turning on of the image commences in the other eye; so similarly it finishes. With the speeds of revolution used for the observations the time the shutter takes to expose or occlude completely each bright disc, varies between .011'' and .002'^ Error that might arise on this score is avoided by the consensual direction of move- ment of the right and left hand shutters.

That the " retinal points " to which the images are thus ap- plied synchronously or in desired sequence are truly " identical " is certified, (i) by the paired physical images being seen single; (2) by the maximum disparation of the edges of the rotating shutters being about 7^ on the retina, whereas 350/^ is about the vertical retinal disparation which Hmits binocular combi- nation. Moreover, a contour travelling through a visual angle of 2° in -^", as in these observations, is not perceptible as a contour at all.

Difficulties due to change in pupil-width are excluded by artificial pupils. Equality of brightness of illumination of the four milk-glass-backed 8mm. holes is obtained by making the straight-wire candle-shaped lamp of considerable, /. e. 12 cm. length, and fixing it accurately in the axis of the cylindrical screen. The rotating screen is blackened inside to minimize reflection. In this way two haploscopic images, one close above the other, are placed in the central field. The right and left components of each of these can be either synchronously or alternately compounded in each. The foveal gaze can be turned from one to the other of them when and as often as the observer desires, and in the fraction of a second by a slight, /'. e. less than 3°, movement of the eyeballs. The comparison thus instituted is facile and sure.

With the apparatus thus arranged various binocular com- binations can be investigated and compared either one with another or with uniocular images. As shown above, the apparatus allows of similar images being thrown on strictly and fully conjugate points of the two retinae, either synchronously right and left or alternately right and left, with a time accuracy not less than .0006" for the slowest rates of intermission, and not less than .0001'' for the highest. The first comparison made (Experiment i) may be to observe if there is any difference between the rates of inter- mittence for just perceptible flicker in two binocular images, one made with synchronous right and left illuminations, the other with alternate right and left illuminations. This arrange-

ment is expressed graphically by the accompanying diagram (Fig. 78). The diagram makes the lower composite image the " syn- chronous " one, but in the series of observations the " synchro- nous " is sometimes the lower, sometimes the upper, and the observer is not informed which it may be. The observations may be made on the transition from flickering to unflickering sensation, or conversely on the transition from unflickering to flickering sensation ; the observer in the latter has a more neutral approach to the critical observation. Compared under rates of intermittence giving marked flicker in both images, observers find the flicker " less " in the " alternate " than in the " synchro- nous " combination. This difference at the lower speeds inclines the observer to expect that complete extinction of the flicker will disappear the more readily in the image which at slow inter- missions seems to flicker the less. Judgment is therefore best asked under conditions in which both images start perfectly free from flicker, the rate of intermittence being from the outset high enough to exclude flicker.

The judgment then given is almost uniformly that there does exist a very small difference between the frequency of intermittence required for extinction of flicker in the " syn- chronous " and " alternate " combinations respectively. In the " alternate *' combination flicker disappears at a slightly lower frequency of intermission than in the " synchronous." All observers agree that directly the frequency of intermission extinguishes flicker in both the discs the appearance of both

is indistinguishably similar, and that there is then nothing to choose between the brightness of the two. For almost all persons I have examined, a spot intermit- tently illuminated at a frequency of intermission just sufficient to extinguish flicker in it, when looked at with one eye only, still flickers slightly when looked at with both eyes. A like phenomenon is noticed by most observers when examined by the arrangement (Experiment 2) represented by Figure 79.

The binocular arrangement, then, is said by them to require a slightly higher frequency for extinction of flicker than does the uniocular. Again, if under a frequency of intermission just securing extinction of flicker in either of the component uniocular images separately, one of these images, previously screened off", is readmitted, so that the pair act together with a synchronous arrangement of phase, a trace of flicker appears at once in the binocular image. It may be urged that this is due to the fresh retinal area being more sensitive to flicker, and it is true that the flicker so introduced tends soon to become less, but a residuum of the phenomenon seems to remain.

Experiment 3. Conversely, under the arrangement indicated in Figure 80, a number of the persons examined, but not all, decide that the binocular image requires for extinction of flicker a slightly lower frequency of intermittent illumination than does the uniocular. Also, a number of these persons, though not all, find that when the " alternate right and left " combination is observed under a frequency of intermission of illumination just sufficient to extinguish its flicker,

the screening out of one of the component uniocular images brings with it a slight appearance of flicker. From these observations it appears that similar phases of flicker- ing illumination if timed to fall coincidently on conjugate retinal areas do very slightly reinforce each other in sensation, and if timed exactly alternately do very slightly mutually reduce. But the broad outcome of the observations is that so far from bright phases at one eye effacing dark phases at the corresponding spot of the other eye, there is hardly a trace of any such interference. To judge from its absence of influence on the flicker rate, the dark phase incident at retinal point A' does not, as regards sensual result, modify the bright phase synchronously inci- dent at the conjugal retinal point A, and conversely. If the brightness of the bright phase or the darkness of the dark phase were lessened at A by A', the rate of frequency of stimulus for extinction of flicker must fall. But except in minute and perhaps equivocal degree it does not alter.

As far as sensual effect goes, the light phases at the one eye practically do not, therefore, interfere or combine with the co- incident dark phases at the other; and conversely. Nor do they, in the alternate left and right arrangement, add themselves as a series of additional stimuli to the like series of stimuli applied at the other eye. If they did the revolution rate of the cylindrical shutter required for extinction of flicker in the upper binocular range LR, Fig. 80, would fall far below that required for extinction in the uniocular. This it does not do. It does not fall at all, apart from the minute difference noted by some persons as mentioned above. A similar result is obtained

under the, in some ways more decisive, conditions (Experiment 4) represented in Fig. 81. With this arrangement no observer in my experiments has ever with certainty detected difference between the uniocular and binocular images in regard to either the apparent rate of the flicker when moderately coarse or the rate of intermission required for flicker extinction. This arrangement (Fig. 81) seems the most crucial for deciding the point. In the " alter- nate right and left " arrangement (Fig. 78, LR, upper com- bination) the instants of change of phase falling together right or left, it might be that it did not matter as regards flicker sensation whether the direction of change was from light to dark or dark to light; the rates of intermission being the same right and left, and the instants of their incidence being synchronous, it might then be that as regards flicker the arrangement was only tantamount to the *' synchronous right and left " arrange- ment (Fig. 78, lower combination) or to the uniocular inter- mittence of the same rate. The arrangement (Fig. 81) avoids this dilemma. Moreover it avoids both the minute reinforce- ment and the minute reduction of flicker inherent, according to the above experience, in the exactly " synchronous " and " alternate " arrangements. It may be termed for conven- ience of reference the " intermediate " arrangement. The physi- ological stimulation it delivers to the conjugate retina is by any mode of count delivered at twice the rate of delivery for either retina considered apart from its fellow. Yet the rate of

revolution of the cylindrical lantern required to extinguish flicker in this experiment remains for the binocular image the same as for the uniocular. There arises the question whether we may regard the dark field covering the area correspondent with that to which in the other retina a bright image is presented, as non-existent visually. That assumption has been made above, and is indicated in the diagrams (Figs. 79, 80, 81). In them, where one image is repre- sented as uniocular, the conjugate area of the other retina is left out of the diagram altogether, as though the latter retina were non-existent, or for the time being blind. This seems per- missible, when care is taken to insure absence of all detail or contour from the dark field presented to the other retina, except for the one component of the compared binocular image. When that field is perfectly void of other contours, and unchanging and borderless, it is found to matter little what depth of darkness it has ; it may be a shade of gray or even a fair white, without perceptibly influencing the sensual vibrations given by the flick- ering image before the other eye. The condition seems com- parable with the familiar disability to see the dark field presented to one closed eye, when with the other eye the observer regards a detailed image.^^ For these reasons the visual image result- ing from the presentation of the bright disc to one eye only, as in the arrangements shown by Figs. 79, 80, 81, was regarded as being a truly uniocular product, uncomplicated by any com- ponent from the other retina. The corresponding area of this latter was considered as for the time being out of action as regards sense, not so much by darkness as by virtue of border- less void homogeneity of field, — as when eye-closure affords visual rest. Under this blankness the " retinal points " become unhitched from the running machinery of consciousness, if — and this is essential — the "corresponding" retinal area be con- currently under stimulation by a defined image.

McDougall's ^32 principle of competition for energy between associate neurones seems at work here, for with both eyes shut the dark blankness of eye-closure does become visible. Even with one eye open, if its field be undetailed and homogeneous, glimpses of the " Eigenschwarz " of a closed eye become obtainable (Purkinje, Volkmann, E. Hering). The accurately converse stimulation of the twin retinal areas might be expected to give some interference of the flicker sen- sations so generated. But the experimental evidence indicates absence (practically entire) of any interference between the flicker processes so initiated. The right and left " corresponding retino-cerebral points " do not when tested by flicker reactions behave as though combined or conjugate to a single mechanism. Their sensual reactions retain individuality as regards time-rela- tions even when completely confluent as judged by reference to visual space.

In the foregoing experiments the flicker sensations of " cor- responding" areas of the two retinae appear (almost entirely) without influence one upon another. But in other experiments the flicker test reveals very considerable mutual influence between reactions initiated at the corresponding areas. Suppose (Experiment 5) two binocular images LR and \p similarly combined from similar uniocular components, all indi- vidually equal in brightness and in intermission frequency. Suppose that of the components of one pair (X/s) one (/)) be replaced (Fig. 82) by an intermittent uniocular image (/)'), of the same physical brightness as that giving the visual image /o, but of considerably higher intermission frequency. In p^ all

licker will disappear at slower speeds of revolution of the lan- tern than those required to extinguish flicker in L or R or \. 'ig. 82 represents the arrangement. The frequency of intermission required to extinguish flicker Xp is then found to be much lower than the frequency required ►r extinction of flicker in LR, or in L or R or \ separately, 'hus the frequency for extinction of flicker in Xp was found observer H. H.) to average 52.2 phases per second as against I1.9 phases per second for LR, or for L, R, or X separately.

Screening image p' out of the binocular combination Xp', when the frequency of intermission was just high enough to free the Xp' image from flicker, at once brought flicker into it ; this disappeared immediately image p' was readmitted to the combination. In this instance the intensity chosen for the steady illumina- tion of the conjugate area was equal to that employed for the uniocular flickering image. The duration of the light phases and the dark per revolution of the lantern was equal, and the light and dark phases of the same intensity in both. But the phenomenon obtains also when the steady uniocular image is less bright (Experiment 6, Fig. 83) or more bright (Experi- ment 7, Fig. 84) than the flickering uniocular with which it is combined. The following example illustrates this.

Experiment 6. In the balanced pair of binocular images LR and Xp made of carefully equalized intermittent uniocular images L, R, X, and p the uniocular image p was replaced by one p' of five times greater rapidity of intermission and giving a steady image of only ^ the bright- ness of the images L, R, X, and p when steady. The frequency of intermis- sion required to extinguish flicker in the binocular image kp' (Fig. 83) was then found to be 72.1 phases per second, whereas in LR, and in A, L and R separately, it was 75.5 phases per second as it had been in the previous. The steady sensation from image p' therefore damped the vibration of the flickering sensation from the conjugate spot under image A. by an amount represented by 3.4 phases per second.

Experiment 7 (Fig. 84) illustrates an observation in which for the image p in the binocular combination \p an image p' was substituted of three times higher frequency of intermission and giving a steady image of one fifth greater brightness than the image L, R, and A. when steady. It was then found that the frequency of intermission required to extin- guish flicker in the binocular image Ap' (Fig. 84) was 57.8 alternate equal phases (of A) per second. Whereas in LR, and in A, L, and R, taken separately, the number of such phases required was 63.6 per second.

The image Xp' was distinctly brighter visually than was LR, or any of the uniocular images A, L, and R. These observations show, as do the observations represented by Fig. 82, that it is not merely the reduction of brightness in the combined image \p in the arrangement shown by Fig. 83 that lessens the flicker in the latter. In fact, the observations on the plan illustrated by Fig. 84 we have the, for flicker photometry, interesting case of a brighter intermittently illuminated surface flickering less than a duller one.

of the steady brightness at one eye to the dark phase of the intermittent at the '* corresponding spot " lighten the latter, and its addition to the phase of equal brightness with it leave that practically unaltered. That might be evidence of mutual interference between purely physiological processes initiated at the corresponding spots of the right and left retinae. But, on the other hand, the result at once suggests that the binocular product from a uni- ocular flickering and a uniocular unflickering image arises by a synthetic process akin to that which produces from a pair of individual uniocular brightnesses a binocular brightness near the arithmetic mean of the brightness of the two components. The rule of combination exemplified by these latter finds little solution by appeal to summation or interference of retinal and purely physiological processes.