J. Physiol. (1978), 283, pp. 223-262 With 1 plate and 17 text-figuree Printed in Great Britain

223

THE PHYSIOLOGICAL EFFECTS OF MONOCULAR DEPRIVATION AND THEIR REVERSAL IN THE MONKEY'S VISUAL CORTEX

By COLIN BLAKEMORE*, L. J. GAREY AND F. VITAL-DURAND From the Iutitut d'Anatomie, University of Lauwanne, Rue du Bugnon 9, 1011 Lauwanne, Switzerland and Laboratoire de Neuropsychologie Experimentale, INSERM U94, 69500 Bron, France

(Received 6 January 1978) SUMMARY

1. 1127 single units were recorded during oblique penetrations in area 17 of one normal, three monocularly deprived and four reverse sutured monkeys. 2. In all animals most cells outside layer IVc were orientation-selective, and preferred orientation usually shifted from cell to cell in a regular progressive sequence. 3. The presence in layer IV c of non-oriented, monocularly driven units, organized in alternating right-eye and left-eye 'stripes' (LeVay, Hubel & Wiesel, 1975) was confirmed. 4. Early monocular deprivation (2-5j weeks) caused a strong shift of ocular dominance towards the non-deprived eye. However, even outside layer IVc, neural background and some isolated cells could still be driven from the deprived eye in regularly spaced, narrow columnar regions. In layer JVc the non-deprived eye's stripes were almost three times wider, on average, than the deprived. 5. Later monocular deprivation (11-16 months) had no detectable influence on layer Ic but seemed to cause a small shift in ocular dominance outside IVc. Deprivation for 61 months in an adult had no such effect. 6. After early reverse suturing (at 51 weeks) the originally deprived eye gained dominance over cells outside layer IVc just as complete as that originally exercised by the eye that was first non-deprived. 7. The later reverse suturing was delayed, the less effective was recapture by the originally deprived eye. Reversal at 8 weeks led to roughly equal numbers of cells being dominated by each eye; fewer cells became dominated by the newly open eye after reverse suturing at 9 weeks and most of them were non-oriented; reversal at 38j weeks had no effect. 8. Binocular cells, though rare in reverse sutured animals, always had very similar preferred orientations in the two eyes. The columnar sequences of preferred orientation were not interrupted at the borders of ocular dominance columns. * Royal Society Locke Research Fellow, on leave from the Physiological Laboratory, University of Cambridge, Cambridge CB2 3EG.

224

C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND 9. Even within layer IVc there was evidence for re-expansion of physiologically determined ocular dominance stripes. After early reverse suture, stripes for the two eyes became roughly equal in width. Possible mechanisms for these changes are discussed. INTRODUCTION

The reorganization of the visual cortex that occurs after monocular deprivation is one of the most striking examples of developmental plasticity in the mammalian brain. Most cells in the visual cortex of very young kittens (Hubel & Wiesel, 1963; Pettigrew, 1974; Blakemore & Van Sluyters, 1975) and newborn monkeys (Hubel & Wiesel, 1977) already receive excitatory input from both eyes; but occlusion of one eye causes the majority of neurones to become monocularly driven by the non-deprived eye (e.g. Wiesel & Hubel, 1963, 1965a; Hubel & Wiesel, 1970; Baker, Grigg & Von Noorden, 1974; Crawford, Blake, Cool & Von Noorden, 1975; Hubel, Wiesel & LeVay, 1977). As little as a day of deprivation can provoke such a change in the kitten (Olson & Freeman, 1975; Movshon & Diirsteler, 1977), as long as it occurs at about 4 weeks of age, the height of a so-called 'sensitive period', which fades away during the third month (Hubel & Wiesel, 1970). The physiological effects of monocular deprivation have been more thoroughly explored in kittens than in monkeys. Yet primates, with the extreme tidiness of their cortical lamination and the orderliness with which afferent fibres terminate there, offer a greater opportunity to discover the structural basis of these effects. It is known that axons from the laminae of the lateral geniculate nucleus serving the individual eyes terminate mainly within layer IV, in a pattern of non-overlapping 'stripes' (see Hubel & Wiesel, 1977). When a micro-electrode passes tangentially through layer IV, it encounters successive groups of neurones strongly dominated by one eye then the other, corresponding to the anatomically defined bands of termination (LeVay, Hubel & Wiesel, 1975). Early monocular enucleation or deprivation leads to shrinkage of the territory occupied by deprived afferent axons (Hubel et al. 1977). Thus, within the fourth layer of the cortex, the physiological effects of monocular deprivation seem to depend on a simple abnormality in the distribution of the two sets of terminals. However, the precise rules by which geniculate axons become sorted out within the cortex of both monocularly deprived and normal baby monkeys are not yet completely clear. It is unknown whether totally different principles underlie the more exaggerated changes that occur in the other layers of the cortex. One technique that has provided a good deal of information for the kitten but has not been used in primates is so-called reverse suturing. Simply reopening the deprived eye (Wiesel & Hubel, 1965b) causes very little re-establishment of input to cortical cells (although a recent study by Mitchell, Cynader & Movshon (1977) did find more convincing recovery after early reopening of the lids). However, if the originally experienced eye is closed at the time that the deprived one is reopened there can be virtually total capture of neurones by the newly opened eye (Blakemore & Van Sluyters, 1974). The effectiveness of this procedure declines the later in the sensitive period it is done; reverse suturing at 4 or 5 weeks of age causes

RE VERSAL OF MONOCULAR DEPRIVATION IN MONKEY 225 rapid, orderly and total re-invasion of input from the deprived eye (Movshon & Blakemore, 1974; Movshon, 1976; Mitchell et al. 1977). The reverse suturing paradigm can shed light on the whole question of the plasticity of input to the visual cortex, on the duration of the sensitive period and on the rules that determine the distribution of afferent axons within layer IV. We have, then, applied the technique in monkeys. A preliminary report of some of the results has already been published (Blakemore, Garey & Vital-Durand, 1978). METHODS

These experiments were performed on eight old world monkeys, one baby and six juveniles, which were artificially reared, and one adult (Table 1). Lid future We used a method of lid suture similar to that described by Blakemore & Van Sluyters (1975). The lid margins were trimmed, under ketamine anaesthesia, and the conjunctiva was dissected from the lids and sutured together over the cornea. The lids themselves, with the tarsal plates intact, were then sutured.

Preparation for recording Animals were premeditated, usually the day before recording, with a corticosteroid (Decadron, 1-5 mg/kg, I.M.) and further doses were given at 24 hr intervals. The monkey was initially anaesthetized with ketamine (20 mg/kg I.M.) for cut-down and cannulation of a superficial vein. Anaesthesia was then maintained during surgery by i.v. administration of barbiturate (Nembutal) or steroid (Althesin) anaesthetic, as necessary. The trachea was cannulated, the head mounted in a stereotaxic instrument after liberal application of topical anaesthetic in the auditory meatus, and the skull exposed in the vicinity of the lunate sulcus. Even in the juvenile monkeys (but not in the adult) it was possible to discern, through the skull, the vascularization marking the sulci. Therefore we were able to choose reliably the positions of penetrations (Text-fig. 1) within area 17, close to the 17/18 border (which is 1-2 mm behind the lunate sulcus) without having to make large exploratory craniotomies. Since even quite small exposures can cause partial herniation of the brain, we usually made a minute craniotomy, 1 mm or so in diameter, and cut a tiny slit in the dura, through which the electrode was brought near the cortex under observation through an operating microscope. A warm solution of agar-agar was then poured around the electrode, filling a well built around the hole with bone wax or dental cement. When the agar had set, the micro-electrode was advanced into the cortex, where stability was remarkably good, even in the superficial layers. The closed eyelids were reopened, the lids retracted with neosynephrine and the pupils dilated with homatropine. The cornea were protected with plano contact lenses and the refractive state, judged by direct ophthalmoscopy, was corrected with additional spectacle lenses for viewing a tangent screen at a distance of 70 cm from the eyes. After paralysis, and at intervals throughout the experiment, the projections of the optic disks and foveae were plotted with a reversible ophthalmoscope on the tangent screen. Artificial pupils were not used because of the difficulty of plotting retinal landmarks through them, especially in young monkeys where the fovea is by no means easily located (Johnson, Tucek & McGowan, 1977). However, in juvenile animals the natural dilated pupil is usually not much more than 5 mm in diameter.

Maintenance, recording and 8timulation The monkeys were paralysed with a loading dose (7 mg/kg i.v.) of Flaxedil and then given an I.v. infusion of Nembutal (1-4mg/kg.hr) and Flaxedil (usually 8mg/kg.hr) in Ringer solution containing 10% glucose, at 3 ml./hr. They were artificially ventilated with air and the tidal volume adjusted to give an end-tidal CO of about 4-5 % (the level maintained by a lightly anaesthetized young monkey breathing spontaneously). We monitored e.c.g., e.e.g. and 8

PHY

283

226

C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND "a

10

eq

q4

eq

eq

aq

eq

-

-

-

r

--"

44,

0

0

..A

h 0X

10

0 00 CV

°

0

ct e

eq

_

0

44 0

o

z4

oo0 0

Po

s

0

10

e.|. 0

D

l 'o

14

\~~~~~~r4U

a

e ut

10to0

0

0

cor

o -o o

0

10 0 CC co

- -

10

10

10

-

w~

0~~ 4Q_ to

0

0

0

0> A 4. e

4j

-4-

0

'p

0 4)

S

CD

s

e

CD

?s

54

0ou ID

0

0

00

pb

*E

R

Ao P

Q

0

0 -

eq

4Q

-

-.4

ct-

co

'p 4> .p .t *4

O~~~~~~~O 4

>

o

>

0

.k~ ~ Po

eo

o o> C*

0 0

co eq

t" t'. 1

~--rz4 P4

00 0

P.4

te p.4

t-. P4

tr

..w

REVERSAL OF MONOCULAR DEPRIVATION IN MONKEY

227

rectal temperature (which was maintained at 38 'C). The e.e.g. gave us a good indication that the infusion of Nembutal maintained a state of light anaesthesia. In a number of animals we omitted relaxant from the infusion mixture, at intervals throughout the experiment, until the monkey could maintain its own respiration, and we then confirmed that it was lightly but adequately anaesthetized, with a little spontaneous movement but only a weak withdrawal reflex. The dose of Flaxedil was adequate to restrict eye drift to a total range of just a few degrees over the whole experiment. All the animals were kept in apparently excellent physiological condition by these procedures and recording was continued for 36-48 hr, during which time we were able to collect an average of 141 units from each animal. Right

-6

Left

\~~~~~~~~~~~

Text-fig. 1. Side views of the brain of a juvenile patas monkey. The dotted line behind the lunate sulcus is the approximate position of the border between areas 17 and 18, which represents the vertical meridian. The filled circles indicate the positions of all penetrations in this study. The arrow shows the standard angle of penetration projected onto the parasagittal plane but in two cases the electrode was exactly vertical. Lines marked + 10 to - 60 plot the approximate projections of horizontal meridians in the visual field (negative = below projection of fovea). Recordings were made with glass-coated tungsten micro-electrodes (Merrill & Ainsworth, 1972). Their uninsulated tips, which were about 15 4am long and tapered from about 3 jsm to less than 1 jam in diameter, were coated with gold and platinum black. Such electrodes recorded rich background activity as well as a high yield of well-isolated single units, even amongst the very small cell bodies in layer IVc. Once a unit had been isolated, the amplified signal was monitored on a storage oscilloscope to check constancy of the wave form. Responses were judged by listening to the spikes on an audiomonitor. The electrode was advanced by a steppingmotor microdrive (Narishige SM-21) with a 1 #sm step size. In combination with finely tapered electrodes, this system has negligible mechanical hysteresis. We were usually able to record excellent units as we withdrew the electrode from each penetration, and found that distinctive features, such as the abrupt transitions of ocular dominance in layer lvc, were encountered at almost precisely the same depths as on the way down. Visual stimuli, such as moving and flashing spots, bars and edges, were back-projected on the tangent screen by manipulating cut-out shapes on the stage of an overhead projector. The dark parts of the pattern had a luminance of about 5 cd/M2 while the light parts were 05-1-0 log unit brighter. A partial reflector between projector and screen cast images on sheets of paper, where receptive fields were plotted. If white light was ineffective in driving a unit, broad-band filters were used to provide coloured stimuli. We only resorted to coloured patterns when units would not respond reliably to white light, so we certainly underestimated the true proportion of colour-sensitive units (Hubel & Wiesel, 1968; Michael, 1972; Dow & Gouras, 1973; Dow, 1974, Yates, 1974; Gouras, 1974; Poggio, Baker, Mansfield, Sillito & Grigg, 1975; Bertulis, Guld & Lennox-Buchtal, 1977). They were indeed few in number but, in common with other investigators, we found them in distinct clusters, perhaps because of a system of colour-selective 'nests' within the cortex (Hubel & Wiesel, 1968). 8-2

228

C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND

Reconstruction of penetration In those penetrations in which the electrode did not reach white matter we made electrolytic lesions at intervals along the track (up to 5 ,AA d.c. for up to 5 sec). At the end of the experiment the animal was deeply anaesthetized with Nembutal and perfused through the heart with heparinized saline, followed by 10 % formalin or 1 % paraformaldehyde/1 -25 % glutaraldehyde in 0-1 M-phosphate buffer. The calvarium was removed and the brain post-fixed overnight. The head was then replaced in the stereotaxic instrument and a scalpel blade mounted in the manipulator, which was set to the same angle as when it held the micro-electrode. Small blocks containing each penetration were then cut out. First, two parallel scalpel cuts were made in the cortex, a few millimetres apart, in planes that were judged to be normal to the surface (roughly parallel to the lunate sulcus). The scalpel blade was then rotated to an angle orthogonal to this first one, and another pair of cuts was made, medial and lateral to the penetration. Each block was undercut and removed, post-fixed and left to sink in a solution of 10% sucrose in fixative. It was then frozen and sectioned at 40 jAm, cutting parallel to the fir8t pair of scalpel cuts. The sections were mounted and stained with cresyl violet. Since the scalpel blade was always parallel to the penetration, this produced sections of the cortex, approximately normal to the surface, in which the two sides of each section (made by the second pair of scalpel cuts) ran across the cortex at the same angle as the electrode track, providing a helpful clue to its orientation. This method made it possible for us to identify every penetration, even those in which lesions were not made. In fact in the majority of cases we found it unnecessary to make lesions: we simply noted the microdrive readings when the electrode first touched the cortex (appearance of background 'swish') and when it reached the white matter (loss of orientation-selective swish and injury discharges and appearance of fibre-like activity for at least a few hundred csm beyond). Since the cortex is relatively flat in the region just above the foveal representation, almost all penetrations, even though quite oblique, did reach white matter. Examination of the sections allowed determination of shrinkage (usually about 5 %) by comparison of the total track distance within the grey matter on the section and the difference between the microdrive readings for surface and white matter. In this paper we use Brodmann's (1909) numbering scheme for cortical lamination, illustrated in PI. 1, which also shows an electrode track.

Experimental strategy for analysis of ocular dominance In order to study the pattern of ocular dominance columns it was essential: (1) to penetrate obliquely, so that the electrode would cross a number of stripes of afferent termination within layer W~c, and (2) to choose a site and orientation of penetration such that the vector of the angle of the electrode parallel to the cortical surface should run roughly orthogonal to the overall pattern of stripes. This avoided intersecting the stripes at a large angle and hence obtaining aberrant estimates of their widths. We therefore made all penetrations in area 17 just above the foveal representation and a few millimetres behind the 17/18 border. Here the stripes run as fairly regular bands, parallel to the representations of horizontal meridians (Text-fig. 1), towards the 17/18 border (LeVay et al. 1975). The electrode was driven slightly antero-medially; the filled arrows in Text-fig. 1 show the angle of penetration projected on the parasagittal plane, running approximately orthogonal to the expected array of ocular dominance stripes. The track usually made an angle of 35-40O to the cortical surface in the plane orthogonal to the surface. In a few cases we used an even shallower angle to provide more nearly tangential penetrations through layer Wc. In order to gain truly representative samples for analysis of ocular dominance we used one of two strategies: (1) studying every neurone that could be isolated or, (2) taking units at regular intervals (usually 100 #tm) by moving the electrode back and forth for a short distance around each stopping place until a cell was isolated (Stryker & Sherk, 1975). In general only one of these strategies was employed throughout a whole penetration. However, large steps were sometimes too coarse to allow precise determination of column boundaries within layer IVc; so in such penetrations we usually switched from regular intervals to continuous searching as soon as the electrode entered IVc (a very distinctive event; vide infra).

REVERSAL OF MONOCULAR DEPRIVATION IN MONKEY

229

One experimenter judged the ocular dominance group of every unit, and he was unaware of the history of each animal until after the experiment. Text-fig. 2 illustrates results for an animal in which two parallel penetrations were made about 1.5 mm apart in the right hemisphere. (This monkey was an adult, about 4 years old, which was monocularly deprived in the right eye for 6j months. However, as we shall demonstrate later, the cortex was apparently unaffected by this procedure and hence can be regarded as normal.) Both tracks extended from surface to white matter (though in penetration R2 units were only sampled down to a depth of 15 mm and then the electrode was deliberately advanced straight to the white matter). On the left of Text-fig. 2 are schematic reconstructions of these two penetrations showing the ocular dominance groups (Hubel & Wiesel, 1962) of units encountered. Dominance meandered back and forth in the manner characteristic of oblique penetrations across the ocular dominance 'columns' (Hubel & Wiesel, 1977). It is almost certain that such 'columnar' regions of relative dominance by one eye then the other are radial extensions of influence, up and down through the cortex, from the totally segregated bands of termination in layer IV (Sokoloff, 1975). Both sequences of ocular dominance moved into a region of strong ipsilateral influence about 0 5 mm from the surface, then into a contralaterally dominated zone at about 1 0 mm and back into an ipsilateral region around 1-5 num or so. In this case the average repeat distance, for transition of a complete left-eye/right-eye 'hypercolumn' (see Hubel & Wiesel, 1977) was perhaps slightly more than 1 mm. When corrected for the angle of about 600 that the tracks made with the laminate through most of the depth of the cortex, this gives a surface-parallel repeat distance of 0-5 mm or more; this is within the range found by Hubel & Wiesel (1972) for the combined width of a pair of stripes of termination, in layer IV. We therefore conclude that the surface-parallel vector of these two penetrations ran approximately orthogonal to the pattern of stripes, and that they probably traversed the same set of columns running forward towards the 17/18 border. RESULTS

Our principal aim was to study the effects of reverse suturing (four animals) but we also recorded from three straightforward monocularly deprived animals and one normal control. We were especially keen to gain information about the sizes of the stripes of termination in layer IV c. Because baby monkeys are scarce we studied large numbers of units from both hemispheres in each monkey: 1127 isolated neurones were recorded in area 17, the minimum number from a single monkey being 72, the maximum 229. All receptive fields were centred between o and 3.50 from the fovea, and most were within 20. They usually lay just below the projection of the fovea, and slightly to the contralateral side of the vertical meridian, in accordance with the positions of penetration (Text-fig. 1).

Part I. Ob8ervation8 on normal cortex We recorded 229 single units from four penetrations in a normal 8-month-old patas monkey (P7711). From this (and from the responses through the non-deprived eye in the monocularly deprived animals) we gained some impression of the organization of the normal visual cortex. For each cell that was selective for orientation, we determined, in each eye, the optimal orientation and the total angular range over which the cell would respond. By testing the responses to stationary bars and edges, and employing Hubel & Wiesel's (1962, 1968) criteria, we attempted to classify each oriented cell as simple, complex or hypercomplex, though in the time available (given that our principal interest was in ocular dominance) this was not possible for every cell. In general our observations confirmed closely those of Hubel & Wiesel (1968, 1977).

230 C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND Simple cells in the vicinity of the fovea often had very small receptive fields, with central summating regions just a few min of arc across. For most of them the central region of the receptive field gave ON responses. They were often selectively sensitive to slow velocities of movement, just as in the cat (Movshon, 1975). We agree with 1

Surface 0

cm

Ocular dominance 1 2 3 4 5 6 7 1 2 3 4 5 6 7 .

:I

0-5

EE 0

.*e, 1* 100

IVi

c

.

C ab U

c

1.5

2-0 White matter

.

Left * Right Penetration R 1

l

I

Left Right Penetration R2

IVc

mm

Text-fig. 2. Reconstructions of two penetrations in the right hemisphere of monkey P62, a young adult patas monocularly deprived in the right eye for 6* months. The lateral view of the occipital pole shows the sites of entry of the two tracks. The diagrams below show two sections, cut in a plane approximately parallel to the lunate sulcus, as described in the text. Laminae I and IVc are indicated, together with the lower boundary of layer VI, and the white matter (WM). The more anterior section, containing penetration R2, passed through the 17/18 border (small arrow). Each penetration is drawn with a thick line (indicated by a filled arrow) running from surface to white matter. The scale below the sections indicates in vivo dimensions, since it is corrected for shrinkage. The two diagrams on the left show the ocular dominance groups (Hubel & Wiesel, 1962) of the cells encountered in these penetrations. Cells in groups 1 and 7 are monocularly driven by contralateral and ipsilateral eyes respectively. Those in groups 2 and 3 are binocularly driven but are, respectively, strongly and slightly dominated by the contralateral eye. Similarly, groups 5 and 6 are slightly and strongly dominated by the ipsilateral eye. Thus the 1-7 scale represents a spectrum of dominance from left eye to right eye in this case, as shown at the bottom of the graphs. The ordinates are the distances along the track. Each point plots the position and ocular dominance of an isolated cell, filled circles being orientation-selective units while open circles are non-oriented units. In the latter part of R2, where the ordinate is interrupted, the electrode was deliberately advanced to the white matter without recording more cells.

REVERSAL OF MONOCULAR DEPRIVATION IN MONKEY 231 Dow (1974), Poggio et al. (1975) and Schiller, Finlay & Volman (1976a) that simple cells can be recorded in most, if not all cortical layers. However, we also concur with Hubel & Wiesel (1968) that a band of distinctive, monocularly driven simple cells is usually encountered just above the non-oriented units of layer IVc. Complex cells were also found in all layers but were very rare in IVc. They seemed less distinctively different from simple cells in their velocity preferences than Movshon (1975) has reported for the cat, but without quantitative methods it is difficult to make this point reliably. A number of cells that were otherwise complex had pronounced direction selectivity even for small moving spots and were very broadly 'tuned'. They were presumably the 'class IV, direction selective complex cells' of Dow (1974) and were similar to the so-called 'pure direction selective cells' in cat cortex (e.g. Palmer & Rosenquist, 1974; Blakemore & Van Sluyters, 1975). We found them quite often below layer IV and it is therefore likely that some of them were corticotectal cells (Finlay, Schiller & Volman, 1976). We came across several hypercomplex cells (Hubel & Wiesel, 1968), especially in the superficial layers (Gouras, 1974; Poggio et al. 1975; Schiller et al. 1976 a).

Non-oriented cells outside layer IVc Hubel & Wiesel (1968), who first discovered units with circularly symmetrical receptive fields in layer IV, have not reported them in significant numbers in other layers. However, other groups have found non-oriented units outside layer IV (Spinelli, Pribram & Bridgeman, 1970; Dow & Gouras, 1973; Dow, 1974; Gouras, 1974; Poggio et al. 1975; Schiller et al. 1976a). We were particularly concerned about this question, because in the normal cat non-oriented cells are very rare, but they are more common after monocular deprivation (Wiesel & Hubel, 1963) and reverse suturing (Blakemore & Van Sluyters, 1974; Movshon, 1976). In the normal monkey we too found a few non-oriented neurones (four out of 114 units) outside histologically defined IVc: by all the usual criteria they were recordings from cell bodies, and three of them were binocularly driven. Sequences of preferred orientation Virtually every penetration (even in the deprived animals) revealed remarkable, progressive sequences of preferred orientation, from one unit to the next, as described by Hubel & Wiesel (1974). Text-fig. 3 contains an analysis of one penetration (L2) in the normal monkey. On the right is a camera lucida tracing of a section containing the penetration, which is drawn as a thick line, indicated by a filled arrow. The positions of two electrolytic lesions are marked. Note the way in which the trimmed edges of the section cut through the cortex at roughly the same angle as the electrode (see Methods). The borders of lamina I, lamina IV c (including both a and fi subdivisions) and the edge of the white matter are also indicated. In this case, exceptionally, the plane of section ran rather obliquely through layer IV, because the penetration was near a dimple in the cortex. The two graphs plot, as a function of distance from the surface, the ocular dominance of the units (on the left) and their preferred orientations (on the right). In the diagram showing ocular dominance, filled circles refer to orientation-selective

232 C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND neurones while unfilled circles indicate non-oriented units (thought to be recordings from cell bodies). Most of the latter were clustered as a dense group of monocular units (groups 1 and 7) in a region of the track virtually identical to histologically defined layer IV c. In the first part of this penetration units were sampled in unusually large steps of 250 ,um. Orientation (degrees) 90 -30 +30 90 -30 -60 0 +60 -60 0 Ia 1

Ocular

dominance 1234567

0

-

0-5 0

1i0-I. EE c

0

1*50

._

w

0o F 0~~

0

0

c C ._

2-0-

0

I

C

^8

In

2-5- I 0 o0 0

3*0-

.

-J

0. so 0

0

o I

0

S

3-5- :

I

, I

@~~-

- -

0.

0

.*

00

Right

Left

\

-

-

/I

-

Text-fig. 3. For legend

see

facing page.

In the graph of preferred orientations, units dominated by the right eye have the preferred orientation of the receptive field (in that eye) plotted as an unfilled circle. Filled circles show the best angle for units dominated by the left eye. Despite the large steps between units in the upper part of the track, there was a fairly clear progression of orientation amongst this sample of cells. Sequences continued without obvious disturbance as dominance switched from one eye to the other. Occasional reversals occurred in the direction of sequence, and the slope of the function relating orientation to distance varied from place to place (and even more so from penetration to penetration), as one would expect from the twisting form of the orientation 'slabs' (Hubel, Wiesel & Stryker, 1977). Text-fig. 5 shows

REVERSAL OF MONOCULAR DEPRIVATION IN MONKEY 233 a particularly impressive example of an orientation sequence that moves through more than 3600 in about 1-5 mm of the track, i.e. about 1-1 mm across the surface. These values are very similar to the maximum length of uninterrupted progression and the maximum rate of shift reported by Hubel & Wiesel (1974). In our total sample of 689 orientation-selective cells from all animals there was a slight anisotropy in the distribution of optimal orientations, preferences close to horizontal and vertical being more common than those for diagonal, though our data were not so striking in this respect as those of Mansfield (1974). Because of the problems of interpreting the true distribution of points when plotting on a repetitive, cyclical co-ordinate, such as orientation (Blaklemore, 1978), we adopted a strict convention when preparing graphs of the type used by Hubel & Wiesel (1974). The preferred orientation of each unit is plotted at the nearest possible point on the orientation axis to the value of the immediately preceding cell. This avoids the problem of deciding arbitrarily the relative positions of points when two oriented units are recorded some distance apart (e.g. at the beginning and the end of layer Ic), and hence where the shift in orientation is considerable. Some caution is needed in the use of such graphs because even random sequences of angle can be made to appear quite regular. However, as shown by Hubel & Wiesel (1974), if units in the monkey are recorded with fairly short gaps between them there are unequivocal sequences with small differences in orientation from one cell to the next.

Layer IVc The distinctive properties of layer IVc (Hubel & Wiesel, 1968, 1977) were most impressive. Typically, approximately half way along each track, the electrode suddenly moved into this layer, where background and unit activity lacked orientation selectivity and were driven entirely monocularly, ocular dominance Text-fig. 3. Analysis of penetration L2 in a normal juvenile patas, P7711. The diagram on the right, as in Text-fig. 2, is a reconstruction of the penetration. In this case the track lay near a small fold in the cortex and the plane of section was not normal to the surface. Consequently the thickness of the layers is exaggerated. This penetration did not reach white matter and therefore shrinkage was calculated from the separation of two lesions, shown as filled circles on the track. The obliquity of the plane of section leads to the angle between the track and the surface appearing larger than it truly was in the surface-normal plane. The two schematic reconstructions show ocular dominance and preferred orientation as a function of distance along the penetration. The ocular dominance scale runs from right to left eye (since the track is in the left hemisphere). Filled circles show oriented units, unfilled non-oriented. The filled bars along the ordinate indicate regions where unresolved background activity could be driven through the contralateral eye (on the left) and ipsilateral eye (on the right). The background and most isolated cells were binocularly driven, except within layer IVc, whose boundaries are indicated by horizontal lines. Inside IVc most cells were non-oriented and monocular, with regular switches in ocular dominance. In the second diagram the preferred angles, in the dominant eye, are plotted for all orientation-selective cells. Filled circles indicate units dominated by the left eye and unfilled by the right. For group 4 cells, the preferred orientation plotted is, arbitrarily, that of the contralateral eye. Down to layer IVc, units were sampled at large intervals of 250 /sm. Nevertheless, a regular sequence of preferred orientations is visible. On the scale of orientation horizontal is zero, anticlockwise angles are positive and clockwise are negative (compare the scale with the line segments below the diagram). By convention, as described in the text, each point is plotted on the abscissa at the nearest position to the preceding point.

234 C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND alternating from region to region. Binocular units were very rare and were nearly always found during a dominance switch. Also, background activity could be elicited through both eyes over only very short distances, usually less than 50 jm, during the transition from one stripe to the next. The units isolated in TV c almost all had circularly symmetrical receptive fields. There is still debate about whether such units are really cortical cells or are merely afferent geniculate axons, many of which terminate in IVc (Hubel & Wiesel, 1972). However, we agree with Hubel & Wiesel (1968) that their receptive fields are often rather different from those of geniculate axons (which we certainly did record from time to time in both grey and white matter, but have entirely excluded from our analysis). Non-oriented cortical units tended to respond less briskly and were more demanding in their stimulus requirements than geniculate units; some of them lacked an obvious inhibitory surround (Dow & Gouras, 1973; Dow, 1974; Gouras, 1974; Poggio et al. 1975). They had typically cellular wave forms quite unlike the tiny, brief action potentials of axons that were sometimes recorded amongst them. They were usually long in duration and negative-going, but sometimes, as the electrode came very close to the unit, the action potential became initially positive, and sometimes tri-phasic, typical of a recording from a cell body (see Text-fig. 6). Further advance of the electrode occasionally caused a prolonged cell-type injury discharge; but perhaps because of the very small size of the stellate cells in IVc, injuries were less common there than in the rest of the cortex. We agree with Hubel & Wiesel (1968, 1977), Dow & Gouras (1973), Dow (1974), Gouras (1974) and Poggio et al. (1975) that there probably is a distinctive population of nonoriented cortical cells in layer IVc, and we have therefore separated them from the rest of the sample in analyses of ocular dominance. Histological reconstruction of each penetration showed close correspondence between the borders of layer IVc and the zone of non-oriented, predominantly monocular units. They never disagreed by more than 100 jsm. In Text-fig. 3, for instance, the boundaries of IVc determined from the histology, are shown on the two schematic reconstructions, where they delimit almost exactly the region of monocular, non-oriented units. Knowing the angle of the track to the cortical laminae, and assuming, as explained above, that the surface-parallel component ran roughly orthogonal to the stripes of axonal termination, we were able to calculate the tangential width of each ocular dominance band within layer IV c. In this way videe infra) an attempt could be made to assess the variability in stripe width for each eye and to determine any consistent difference in width for the two eyes. Most of our penetrations only stayed in JVc long enough to pass across about three switches in dominance, but some were more productive. The long track (L2) in the normal animal (Text-fig. 3) remained in layer IVc for more than 5 mm. It crossed five switches in ocular dominance, and there were no consistent differences in size between regions devoted to the right eye (group 1) and those to the left (group 7). Orientation-selective cells, usually simple and often monocular or very strongly dominated by one eye, were sometimes, as in Text-fig. 3, recorded within the boundaries of IVc, mixed in with non-oriented units.

REVERSAL OF MONOCULAR DEPRIVATION IN MONKEY

235

Ocular dominance The sample of units was always divided into two classes when considering their ocular dominance: (1) all non-oriented units with cellular wave forms, within the histological boundaries of layer IV c plus 100 /tm on either side, to allow for a margin of error in reconstruction, and (2) all other non-oriented cells, together with all orientation-selective (and direction-selective) cells wherever they were recorded. Non-layer lVc

10

-

-

0-

1

3

2

4

5

6

7

2

1

3 4

5

6

7

10

-0

Layer IV c

40 -

- 40

20 -

- 20

-0

01

2

3

Right

45

6

Left

Left hemisphere

7

1

2

3

4

5

6

7

Right Right hemisphere

Left

.

Text-fig. 4. Histograms for the ocular dominance of units recorded in the left hemisphere (on the left) and the right hemisphere (on the right) in the normal animal, P7711. The ordinates plot the numbers of cells. The upper histograms are for all orientationselective cells (filled blocks) and all non-oriented cells recorded more than 100 tm outside the histologically determined boundaries of layer IVc (unfilled blocks). The lower histograms are for all non-oriented cells recorded within the boundaries of IVc, plus 100 pum on each side.

Text-fig. 4 plots conventional ocular dominance histograms, separately for the two hemispheres, and for the two classes described above, for the normal juvenile

monkey. Filled blocks indicate units that were orientation selective through their dominant eye, and unfilled blocks represent non-oriented neurones. The distribution

236 C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND for the IVc non-oriented cells is, of course, highly monocular. The apparent overall dominance (by the right eye in the right hemisphere and the left eye in the left) is almost certainly a sampling error: because of the fact that only a small number of stripes are traversed in a single experiment, there is a high probability that the electrode will spend more time in areas dominated by one eye, even if the stripes are identical in width. The ocular dominance for 'non-layer IVc' units is similar to Hubel & Wiesel's (1968) distribution for neurones outside IVc (though we found more units in the middle groups, as did Schiller et al. 1976b). Interestingly, there appears to be no pronounced overall dominance of the monkey's cortex by the contralateral eye, unlike the situation in the cat (Hubel & Wiesel, 1962; Blakemore & Pettigrew, 1970). Ninety-eight of the 114 'non layer IVc' cells (86 %) were binocularly driven. Part II. Monocular deprivation We recorded from three animals that had been monocularly deprived in the right eye (Table 1). The deprived eye was not reopened until the start of the recording.

Early deprivation One vervet monkey (V7710) was deprived in the right eye from 2 to 51 weeks of age, before recordings were taken. The result was a marked shift in dominance towards the left eye. The reconstruction of a penetration (LI) in the left hemisphere (Text-fig. 5) shows that all but six of the seventy-two orientation-selective cells recorded in this track were dominated by the left eye and the majority of units were in group 7. Close to the upper border of layer IVc there was a sudden appearance of monocular non-oriented cells, all of which were dominated by the left eye for about 600 jum. Then, as the electrode was leaving layer IV, there was a tiny region, about 100 ,um long, where background activity and units were exclusively driven by the deprived right eye and two non-oriented units were isolated. Despite the virtually ubiquitous rule of the left eye, there was an impressive tendency for influence from the deposed right eye to linger on in a highly regular ' columnar' pattern, even outside layer IV c. In Text-fig. 5, the filled bars along the edge of the ocular dominance graph show the regions in which unresolved background activity could be influenced through the contralateral, right eye (on the left) and the ipsilateral, left eye (on the right). There were only two small regions, including the patch already described at the edge of layer IVc, where there was no such activity driven from the non-deprived, left eye. But there was background 'swish' from the deprived eye too, in restricted areas about 100 #um across, at about 500 ,um intervals throughout the track. Within these areas there were almost always some cells with a receptive field in the deprived eye, and in some cases (see the end of the track) units were monocularly driven by the deprived eye. This persistent, periodic representation of the deprived input, even after a significant episode of deprivation, is rather different from the situation in the cat, where background activity from the deprived eye (at least outside layer IV) is rather uncommon (Wiesel & Hubel, 1963, 1965a). It is hard to avoid concluding

REVERSAL OF MONOCULAR DEPRIVATION IN MONKEY 237 that the narrow regions of influence from the deprived eye throughout the cortex represent radial extensions of activity relayed from the surviving shrunken stripes of termination from the deprived afferent axons, found in layer IV (Hubel et al. 1977).

Note that in the reconstruction of the sequence of preferred orientations, in Text-fig. 5, the few cells dominated by the deprived eye (open circles) had preferred orientations that fitted into the over-all progression for the other eye. A number of Ocular dominance 1 2 3 4 56 7

Orientation (degrees) 90

+30

-30

0

-60

-90

90

+30

-30

+60 -60

-30

+60 -60

0

0

~ 0-5-

~

~

~

1-0~ ~ ~ ~ E~~~~~~~~~~~~ C~~~~~~~~

0~~~~~~~~~~~~~~~~~~~00

Co 0

08

C X 20__

a

00 .

0

*0

25-1 .*l

0.: 0

lS

~~~I

ns

.00

*

303-0-

____I_

Right

Left

X

Text-fig. 5. Reconstruction, as in Text-fig. 3, of penetration LI in V7710, deprived in the right eye from 2-5i weeks of age. The sequence of preferred orientations was particularly striking in the first part of the track, moving through a total range of more than 3600 at a surface-parallel rate of about 360'/mm. One group 4 unit, recorded at 1-25 mm, although orientation-selective in the left (ipsilateral) eye, was nonoriented in the right. Therefore the preferred orientation plotted is that of the left eye.

binocularly driven cells still had clear orientation selectivity in the deprived eye (though there was a tendency for the tuning to be broader and the responses more variable). In this animal, out of twenty-seven binocular cells that were orientation selective through the normal eye, ten were also selective through the deprived eye. The remainder had circularly symmetrical fields in the deprived eye, as is often reported for the few remaining binocular units in monocularly deprived kittens (Wiesel & Hubel, 1963) and monkeys (Crawford et al. 1975).

C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND

238

Later deprivation There is little published evidence about the termination of the sensitive period for monocular occlusion in the monkey. The behavioral work of Von Noorden and his colleagues (see Von Noorden, 1973) suggests that a detectable and persistent deficit in acuity for the deprived eye only occurs with occlusion before about 2 months Ocular dominance 1 2 3 4 5 6 7

E

E 1-

I.

5mV

/

v

0

4

a) C I) 0.

l

1 msec

c

I 0

2 mVI

4

1

mm

i

Left

Right

Text-fig. 6. The ocular dominance sequence for penetration RI in F7603, a cynomolgus deprived in the right eye from 11-16 months. The inset photographs on the left show two action potentials (positive upward) of non-oriented units, presumed to be cells, recorded in layer WVc at the positions indicated by the arrows.

of age. And it is certain that monocular closure, lasting more than a few days, between birth and the second month has pronounced effects on cortical ocular dominance (Hubel & Wiesel, 1970, 1977; Baker et al. 1974; Crawford et al. 1975).

RE VERSAL OF MONOCULAR DEPRIVATION IN MONKEY 239 In view of Hubel & Wiesel's (1977) suggestion that deprivation starting after about 6 weeks does not influence the stripes of termination in layer IV but can, if prolonged, produce other defects, we looked at one cynomolgus monkey (F7603) whose right eye was closed from 11-16 months of age. Non-layer IVc

m

0

- 10

10-

-0

0-

12

3

4

5

6

2 3

1

7

4

5

6

7

Layer I V c

10

10

-

II

0 -

12 3 4

5

6

Left Left hemisphere

R ight

o-

7

1

-0 2

34

567

Left Right Right hemisphere -

Text-fig. 7. Ocular dominance histograms for the samples from the two hemispheres of F7603, as in Text-fig. 4. There is slight overall domination by the left (non-deprived) eye in both hemispheres for 'non layer IVc' units. In layer W~c itself, however, there is no such trend.

Our first impression was that there was nothing unusual about the ocular domidistribution in this animal. The reconstruction of penetration RI, in Text-fig. 6, illustrates the rich yield of binocularly driven units outside layer IV. And within IVc there was no evidence that left-eye regions were wider than those of the right eye. Close examination of the data from outside IVc, however, suggests a slight shift in favour of the contralateral left eye. Indeed, in two small regions outside layer IVc background activity could not even be detected from the deprived right eye. An unusually large number of cells was monocularly driven by the normal eye. The ocular dominance histograms (Text-fig. 7) show that this trend in favour of the left eye was present in both hemispheres. Our conclusion is that late deprivation, even at the end of the first year, if

nance

240 C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND prolonged, can probably cause a small shift in ocular dominance for neurones outside layer IVc. It seems, however, to have no impact whatever on the stripes of IVc. Monocular deprivation in an adult One adult monkey (P62), some 4 years old, was deprived in the right eye for 61 months. In brief, its visual cortex was not detectably different from normal: the binocularity, responsiveness and orientation selectivity were well within the range of normal variation. In particular, of the sixty-nine orientational units recorded, only five (7 %) were monocularly driven by the non-deprived, left eye. Two penetrations reconstructed in Text-fig. 2 show the normal arrangement of ocular dominance columns outside JVc. In two of the three tracks in this monkey we recorded no non-oriented units at all. The only penetration with a typical interlude in IVc was RI (Text-fig. 2), but because of the unusually steep angle of the electrode the distance traversed was about 0-3 mm and only three non-oriented units were recorded, all dominated by the left eye. We can draw no conclusions about stripe width in this animal; but in view of the failure of deprivation from 11 to 16 months to influence IV c, there is no reason to expect any change in an adult.

Comparison of ocular dominance histograms In order to compare the over-all results in normal and monocularly deprived animals, the ocular dominance histograms for samples from the two hemispheres were combined, as if they were all from the right hemisphere. In other words, in the combined histograms (Text-fig. 8) cells in groups 1, 2 and 3 are totally, strongly or slightly dominated, respectively, by the left eye (regardless of which hemisphere they came from). Cells in groups 5, 6 and 7 are dominated by the right eye. As already explained, data are analysed separately for non-oriented units that came within the defined boundaries of layer IVc (on the left in Text-fig. 8) and all other neurones (on the right). The histograms for the normal animal (derived from Text-fig. 4) vindicate the strategy of sampling: both within IVc and outside it, roughly equal proportions of units were dominated by each eye. For the animal with early monocular deprivation (V7710) there is a suggestion that the proportion of units dominated by the deprived eye remained higher within IVc (17 % in group 7) than outside IVc (10 % in groups 5, 6 and 7). This matches our impression that representation of the right eye persisted as shrunken but otherwise normal 'columns' in layer IV c; but its influence through the rest of the cortex, while periodically distributed in just the same way, was often expressed as no more than a murmur of background activity. Text-fig. 8. Ocular dominance histograms for 'non layer W~c' and 'layer IVW' units (see Text-fig. 4 and text) for the normal animal (P7711) and the three animals monocularly deprived in the right eye (V7710, F7603 and P62). Data from the two hemispheres are pooled and plotted as if they all came from the right hemisphere (group 1 units are dominated by the left eye, group 7 by the right). No data are shown for layer IVc for the monocularly deprived adult, since only three non-oriented cells, all driven by the left eye alone, were recorded in this animal.

RE VERSAL OF MONOCULAR DEPRIVATION IN MONKEY Normal monkey

50 -

- 20

0-

m

I

-

10

-

0

-

100

Monocular deprivation

20 2-51/ weeks 10

- 50

-

0-

-

0

-

20

-

10

11-16 months

1. H

20 -

M

10 -

0-

Left

-

-0

Right

Layer I V c

- 10

Adult for 6% months

Left Rigi Non layer lVc

Text-fig. 8. For legend

see

facing

page.

-0

241

242 C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND The results of occlusion from 11 to 16 months (F7603) support Hubel & Wiesel's (1977) statement that columns within layer IV are not influenced by even prolonged late deprivation. However, the results from outside IVc, derived from Text-fig. 7, do suggest a shift towards the left eye: sixty-seven units, out of 106 (63 %), were in groups 1, 2 and 3. The results from 'non layer IVc' units in the adult deprived for 6j months (P62) are not significantly different from those from the normal control animal. Part III. Reverse suturing Four monkeys were initially deprived in the right eye, starting within 2 days of birth, or in one case at 3 weeks, and were reverse sutured at ages ranging from 51 to 381 weeks (Table 1). It is known that in the cat the consequences of reverse suturing are complete within 2 weeks or so, whatever the age at reversal (Movshon, 1976). So, to be safe, we allowed a period of at least 4 months (from 16 to 211 weeks) for the animals to use their originally deprived eyes before recording.

Early reversal: at 51 weeks The patas (P7708) that was reverse sutured at 51 weeks was a close experimental counterpart to the vervet (V7710), that was simply deprived until 51 weeks and then recorded. It is reasonable to assume that the cortex of P7708 was similar to that of V7710 (see Text-fig. 5) at the time that it was reverse sutured (especially since its right eye had been closed from 2 days of age). How different the situation was when we recorded from P7708, after 16 weeks of use of the initially deprived right eye. The penetration (L2) illustrated in Text-fig. 9 is from the left hemisphere, so it can be compared directly with Text-fig. 5. The two ocular dominance distributions are virtually mirror images: in Text-fig. 9 almost all units are in group 1: monocularly driven by the initially deprived right eye. Outside layer IVc there was only one small region, about 3 0 mm along the track, where there was no background activity from the right eye and cells were monocularly driven by the left eye. Just as the meagre residual influence from the right eye in V7710 was distributed in small regular patches along the track (Text-fig. 5), so, in P7708, the surviving activity from the left eye occurred in periodic regions where background activity and a few isolated units could be driven through the initially experienced left eye. Not only were the majority of cells monocularly driven by the right eye, as if the initial period of deprivation had never happened, but their receptive field properties were normal. In this track, all cells beyond the immediate vicinity of layer IVc were orientation selective, their 'tuning' was not abnormal and the sequence of preferred angles (Text-fig. 9) was quite regular. Note too that the preferred orientations of the two units dominated by the left eye fitted in reasonably well with the over-all sequence of right eye orientations. In the four penetrations in this animal, seventy units (outside IVc) were dominated by the right eye; sixtynine were orientation selective and one was non-oriented through that eye. Only ten units were dominated by the left eye; six were oriented and four non-oriented. One unit, dominated by the right eye, had an oriented receptive field in that eye

REVERSAL OF MONOCULAR DEPRIVATION IN MONKEY 243 but was non-oriented in the left. In addition we found five cells that we could not excite with any of our visual stimuli - something not seen amongst the 640 units recorded in normal and monocularly deprived animals. Ocular dominance 1 2 3 4 5 6 7

Orientation (degrees) 0 +30 +60 90 -60 -30 0

0*5-

1-0-

E E

1-5 -

0 0 +a

m

0_

c

2-0 -

._

CU

C] C._ 0

2-5 1

mm

I1

3*0 -

3.5 -

I. I\

l Text-fig. 9. Reconstruction, as in Text-fig. 3, of penetration L2 in P7708, a patas that was deprived in the right eye until 51 weeks, then reverse sutured and recorded 4 months later. Compare the ocular dominance sequence with that for the matched monocularly deprived animal in Text-fig. 5. --

Reversal at intermediate ages: at 8 and 9 weeks Two patas were deprived in the right eye until 8 weeks (P7705) and 9 weeks (P7703), when they were reverse sutured and recorded 211 and 18 weeks later, respectively. Text-fig. 10 contains a reconstruction of a penetration (R2) from the right hemisphere of the animal reversed at 8 weeks. The pattern of ocular dominance was

244 C. BLAKEMORE, L. J. CAREY AND F. VITAL-DURAND strikingly monocular. Even outside layer IV c there was regular alternation between groups of cells dominated, usually completely, by one eye then the other. In the middle of each 'column' dominated by one eye, there was often not even background activity from the other eye. Roughly equal numbers of units were dominated by each eye, even within layer IVc. Thus the degree of recapture by the initially deprived right eye was less complete than after earlier reverse suturing (cf. Textfig. 9). The sequence of preferred orientations was regular with no discontinuity as dominance switched from one eye to the other.

E E 0

C

am c

w CD

lI

Left-'- Right

I \

/ \ - - -/ \ in of R2 Text-fig. 10. Reconstruction penetration P7705, a patas reverse sutured at 8 weeks of age. - -

Penetration L1 in the animal reverse sutured at 9 weeks (Text-fig. 11) was even more monocular. The newly experienced right eye had been less successful in taking over the cortex and many of the cells monocularly driven by the right eye, even outside layer IV c, were non-oriented. Late reverse suture: at 381 weeks One cynomolgus (F7601) was reverse sutured at 38J weeks. Text-fig. 12 (when compared with the results for the monocularly deprived animal in Text-fig. 5) shows that the cortex was apparently unaffected by the 161 weeks of use of the originally deprived right eye. The cortex was as much dominated by the left eye as that of V7710; in this penetration, influence from the right eye was only found in two regions, one of them in layer IVc.

REVERSAL OF MONOCULAR DEPRIVATION IN MONKEY

245

Comparison of ocular dominance histograms Data from the two hemispheres were combined, as in Text-fig. 8, for all the reverse sutured animals (Text-fig. 13), with non-oriented cells in layer IVc again shown separately. Comparison with the results for the 2-5i week monocular deprivation (Text-fig. 8) shows that reverse suturing at 51 weeks produces recapture Ocular dominance 2 3 4 5 6 7 0

Orientation (degrees)

-_

05 -

1.0 -. EE c

.2 1-5 T

4, 4,1 0.

CL co

-

20 a 2-5

1 mm

3*0 I

\N-- ---/

\

--

Text-fig. 11. Penetration Li for P7703, a patas reverse sutured at 9 weeks. In this track units were sampled at regular intervals of 100 /tm.

by the newly experienced eye just as complete as the original degree of takeover by the initially experienced eye. Late reversal, however, seems not to change cortical ocular dominance at all; and reversal at the end of the second month causes an intermediate degree of recapture. Just as in reverse sutured kittens (Blakemore & Van Sluyters, 1974; Movshon, 1976), binocular neurones are extremely rare. We have not looked at animals with very brief periods of reverse suture; in kittens binocular cells are somewhat more commonly encountered during the dynamic phase of recapture (Movshon & Blakemore, 1974; Movshon, 1976).

246 C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND Another characteristic of reverse sutured animals is the greater than normal proportion of non-oriented units outside layer IVc. This was especially true for units dominated by the originally deprived eye in the animal reverse sutured at 9 weeks. The same thing is seen in kittens reverse sutured at the start of the third month (Blakemore & Van Sluyters, 1974). Equally, the occasional occurrence of visually unresponsive cells was probably significant, since we were able to drive every isolated neurone in the normal and monocularly deprived monkeys. Even amongst the visually responsive units a proportion were unusually sluggish, had poor stimulus selectivity and were rapidly fatigued by repetitive stimulation. This was particularly true for units dominated by the initially deprived eye in the two animals reverse sutured at intermediate ages. Movshon (1976) has shown that the

0-

Ocular dominance 1 2 3 4 5 6 7

0-5-~

~

~

Orientation (degrees) -60-30 0 +30+60 90-60-30 0 +30

*

*

.2

E

iutr

4

Pw ees 0 oh

1

17

REVERSAL OF MONOCULAR DEPRIVATION IN MONKEY 255 Wiesel (1977) to the idea that competitive interaction, with simple displkcement of underactive or relatively sparse terminals, might explain both the segregation into uniform stripes in normal animals and the apparent expansion of the non-deprived stripes at the expense of those of the deprived eye in a monocularly deprived monkey. One bonus from this theory is that it solves part of the mystery of the sensitive period. On this hypothesis, the sensitive period (for layer IV, at least) would be the post-natal time during which the natural sorting out of terminals by competition is taking place. The effects of monocular deprivation would then be merely the bizarre exaggeration of a natural process. The theory thus gives a possible functional role to the sensitive period, but it does not explain its timing, which, like so many examples of the temporal precision of embryological development, remains a mystery. Nor does it account for the extension of the phenomenon and the possible existence of a different sensitive period outside layer IV. However, as a model of what happens in the fourth layer it is most compelling. One corollary of this hypothesis is that axonal 'sprouting' and invasion ofpreviously unoccupied space are not needed to account for the 'expanded' columns of a nondeprived eye. Since axons from both eyes are initially everywhere in layer IV, irreversible displacement of the unsuccessful ones alone might account for the final pattern. If axonal sprouting cannot occur, the stripes of a deprived eye, once caused to shrink by deprivation, should be incapable of re-expansion. This is the reason for our special interest in IVc. Ideally this question must be tackled with anatomical techniques. The method of transneuronal transport of radioactive label (Wiesel, Hubel & Lam, 1974) should be used in reverse sutured monkeys to visualize the stripes and to see whether re-expansion has indeed occurred. But, in the meantime, our data may throw a little light on this issue. We deliberately drove the electrode across the normal pattern of stripes in a region where they are fairly regular. In all but two penetrations the electrode stayed in layer IVc long enough to pass at least one switch in ocular dominance, and often three or more. Thus, knowing the angle of the electrode to the cortical surface, we were able to calculate the surface-parallel widths of the physiological 'columns' within IVc itself. Text-fig. 16 is a schematic reconstruction of the part of each penetration that lay in layer IVc. Every horizontal bar represents one penetration. The left-hand vertical line indicates the start of IV c and the short vertical line on each horizontal bar shows the end of IV c. Within each bar the filled areas represent regions intensely dominated by the left eye; the unfilled areas are right-eye columns. In order to use information at the boundaries of IVc we estimated the upper edge of the first column and the lower edge of the last column by inspecting the columnar pattern of ocular dominance for the surrounding units, outside IVc, assuming that the columns detected in supra- and infra-granular layers are radial extensions of those in IVc. As Hubel & Wiesel (1977) have noted, it is usually possible to specify column boundaries even outside IVc to within 50-100 /sm. In monocularly deprived and reverse sutured animals the situation is even clearer because of the emphasized columnar distribution of background and unit activity. The scale in Text-fig. 16 indicates the surface-parallel distance travelled by the

256

C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND 1

0

.

iiiC *

0 0 4. o

10 O

0 o o

m

la =

10 t'

o X

I I

*

*

0p *

.20 I.C OD

o_ _

D

b

w

&C)

W a

0

0e C)

-

8~~~4

O~4 10 00

CO

-K

CEoo

CO

e

o

4-D 4

e

t U1C0

._q

4-

0

00

00~~~~~~. 0

_ _^a

--

_1

w

(I 0

-6a -4

.

0 0

-2bo9

O

1 .M

*4

g

CD

ceU~t

s

o ec* ,g~ e: S1u>_0t W n X

X0

_

e

eq

ill~

4bW

>

pn X t

C

"

co

c

~10CO0(M

ID I+ +

:e0 0

I

10U'

-4 C.-

C.-

t~.

+I

V-

C

4 0

r-

o+ 0

*4W-o+

ec

co~04

I

*

W

~~4 0~~0

"~

...4

M

0.

C

.-

= p.4

N

C-.

co

C-.

r-

4p4

H

REVERSAL OF MONOCULAR DEPRIVATION IN MONKEY 257 electrode, correcting for its angle of incidence. From these diagrams we measured each column width. The means are listed in Table 2, separately for each eye, together with the sum of the mean stripe widths in the two eyes and the ratio of left-eye to right-eye mean stripe width. The ratio was 2-85 for the monocularly deprived animal, V77 10, very similar to the maximum value of 2%77, derived by Hubel et al. (1977) from direct measurements of anatomically demonstrated stripes in monocularly deprived monkeys. This indicates that the deprived stripes were 'shrunken' by nearly 50 % from half the width of an average pair of stripes. The ratio is similar, 2-72, for F7601, the animal reverse sutured at 381 weeks, implying that the originally deprived stripes remained as shrunken as they were after the initial period of right eye deprivation. On the other hand, the three animals with earlier reverse suture had ratios close to 1D0. Indeed, for P7708 it was 0*67, meaning that right-eye stripes were actually slightly wider on average than left-eye stripes. Given the small samples (4-6 stripes per eye per animal) no single result is compelling, but the over-all impression from these three monkeys (based on a total of thirty individual stripes) is that the originally deprived right-eye columns re-expanded and those of the left eye shrank until they were roughly equal in width. From these data on stripe widths we calculated an induction index (defined as L/(L+ R)) for the normal and monocularly deprived animals, and a reversal index (R/(L+R)) for the reverse sutured monkeys (where R, L are the mean widths of right-eye and left-eye stripes, respectively). These are plotted in Text-fig. 15C with the ordinates again normalized so that the horizontal lines indicate the expected limits of variation. The upper line lies at 0 74 (the induction index for V7710) on both ordinates. This would be the expected reversal index in a reverse sutured animal if the originally deprived columns became as expanded as the originally non-deprived columns used to be. On the left ordinate the lower line is at 0-5 (equal stripe width for the two eyes) and the data for the normal animal (square) and the 11-16 month monocularly deprived animal are close to this value. This analysis, since it rests on calculations of ratios of stripe width, does not depend crucially on the accuracy with which the electrode traversed the stripes orthogonally in the surface-normal plane, nor on the precision with which the plane of section for histological reconstruction fell exactly perpendicular to the cortical surface. In a few cases the plane of section was not orthogonal to the surface, and the cortex appeared in the sections to be significantly thicker than its true value of less than 2 mm (e.g. Text-fig. 3). Any tendency in this direction would lead us to overestimate the angle between the electrode and the surface (in the surface-orthogonal plane), and hence to underestimate the projected surface-parallel width of each stripe. This might account for the fact that our values of about 440-670 pum for the repeat distance of a pair of stripes are somewhat low compared with that of more than 800 /sm obtained by Hubel et al. (1977). On the other hand, there might

be species differences in stripe width: Hubel et al. (1977) used macaques exclusively. If at all times the physiologically determined ocular dominance columns in IVc reflect exactly the pattern of termination of afferent axons, Text-fig. 15C indicates that reverse suturing causes axonal sprouting into initially evacuated territory. In normal monkeys, increased axonal arborization and synaptogenesis may well be occurring in IVc during the first few weeks of life, as the stripes become con9

PHY

283

258 C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND solidated; so perhaps it would not be surprising if axonal growth were still possible after the stripes have been shrunken by deprivation. Additional support for the possibility of axonal sprouting within layer IVc after reverse suturing comes from measurements of the sizes of cell bodies in the lateral Monocular deprivation

Reverse suturing

mm-uN~umuSmU a Ien-

9

*

E

B -

-~ ~-DP I

* RS

D

I

4I if_4

Text-fig. 17. This shows one possible model, discussed in the text, for the changes in the distribution of axons within layer IVc in monocularly deprived and reverse sutured animals. Filled areas indicate regions in which left-eye axons form phy8iologically functional synapses with cortical cells, while the hatched areas show regions within which effective terminals from right-eye axons are present. Unfilled regions show areas in which hypothetical 'synaptic suppression' has occurred, so that terminals, though still present, are incapable of driving cells. At birth, axons from both eyes are distributed throughout layer IVc, though the banding is already visible as a waxing and waning of density (Rakic, 1977). If the right eye is deprived (left-hand diagram), its areas of effective termination might rapidly shrink because of its competitive disadvantage, but it might leave ineffective terminals in left-eye territory for some time. Eventually these silenced terminals might also withdraw, producing the typical close agreement between anatomical and physiological stripes (Hubel et al. 1977). If reverse suture (RS) occurs early enough (right-hand diagram), before degeneration of the postulated 'silenced synapses', the right-eye axons would be put at a competitive advantage in areas of overlap and might ultimately become reactivated to form apparently re-expanded stripes.

geniculate nucleus in these and other monkeys (Vital-Durand, Garey & Blakemore, 1978). Just as in the cat (Dfirsteler, Garey & Movshon, 1976), reverse suture can alter the disparity in cell size in the geniculate laminae caused by monocular deprivation. Early reverse suture corrects or even inverts the size difference between deprived and non-deprived laminate. The initially small deprived geniculate cells become relatively enlarged, to about the same size as cells in the originally experienced laminae, in just the same manner that the stripes belonging to the two eyes become

REVERSAL OF MONOCULAR DEPRIVATION IN MONKEY 259 modified in the cortex. Perhaps perikaryal growth and axonal arborization are linked by some correlated mechanism of growth (Guillery, 1972). These results on re-expansion of fourth layer stripes are preliminary and refer only to physiological column size. In adult monkeys, either normal or after monocular deprivation, there is agreement between the physiologically determined switches in ocular dominance in layer IV and the edges of the anatomically determined stripes (LeVay et al. 1975; Hubel et al. 1977; Hubel & Wiesel, 1977). It is not yet certain, however, that the anatomical redistribution of terminals that can be visualized in a monocularly deprived animal at many months of age has already happened by the age of 5 or 6 weeks, when the physiological shrinkage of the deprived columns is already clear. Could it be that within the intermixed plexus of axons that fills the neonatal layer IV the effect of monocular deprivation is first to cause functional 'suppression' of synapses (see Mark, 1974) in areas where terminals from the other eye have a competitive advantage through both weight of numbers and activity? The deprived input might then only retain influence within apparently shrunken stripes when measured physiologically at just a few weeks of age. Perhaps, though, it takes some time more before the 'suppressed' terminals are functionally withdrawn and the anatomical disparity in distribution becomes clear. During such a transition phase, reverse suturing could put the 'suppressed' terminals at a competitive advantage and enable them to displace the originally non-deprived axons that had settled beyond their normal territory, without axonal sprouting having to occur. This hypothesis, illustrated schematically in Text-fig. 17, would allow the simple competitive theory to cope with the results on reverse sutured monkeys, but only through the introduction of the concept of synaptic 'suppression'. More work is needed to demonstrate the re-expansion incontrovertibly and to discover the true mechanism. This work was supported by grants from the Swiss National Science Foundation (3.2460. 74), the European Training Programme in Brain and Behaviour Research, and the Medical Research Council, London (G972/463/B). C.B. held a fellowship from the Roche Research Foundation for Scientific Exchange and Biomedical Collaboration with Switzerland. Institut M6rieux and IFFA M6rieux, Lyon, kindly donated all the monkeys. We are very grateful to A. Ainsworth, M. C. Cruz, C. Frenois and M. Gaillard for technical help, and to Dr S. M. Zeki for helpful advice. REFERENCES

BAKER, F. H., GRIGG, P. & VON NOORDEN, G. K. (1974). Effects of visual deprivation and strabismus on the response of neurons in the visual cortex of the monkey, including studies on the striate and prestriate cortex in the normal animal. Brain Res. 66, 185-208. BERTuTL, A., GULD, C. & LENNOX-BUCHTAL, M. A. (1977). Spectral and orientation specificity of single cells in foveal striate cortex of the vervet monkey, Cercopithecus aethiop8. J. Physiol. 268, 1-20. BLAKEMORE, C. (1978). Maturation and modification in the developing visual system. In Handbook of Sen8ory Phy8iology, vol. VIII, Perception, ed. HELD, R., LErBowITz H. W. & TEuBER H.-L. Berlin: Springer Verlag (in the Press). BLAKEMORE, C., FIORENTrINI, A. & MAFFEI, L. (1972). A second neural mechanism of binocular depth discrimination. J. Phygiol. 226, 725-749. BLAKEMORE, C., GAREY, L. J. & VITAL-DURAND, F. (1978). Reversal of physiological effects of monocular deprivation in monkeys. J. Physiol. 276, 47-49P. BLAKEMORE, C. & PETTIGREW, J. D. (1970). Eye dominance in the visual cortex. Nature, Lond. 225, 426-429. 9-2

260

C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND

BLAKEMORE, C. & VAN SLuYTERS, R. C. (1974). Reversal of the physiological effects of monocular deprivation in kittens: further evidence for a sensitive period. J. Physiol. 237, 195-216. BLAKEMORE, C. & VAN SLUYTERS, R. C. (1975). Innate and environmental factors in the development of the kitten's visual cortex. J. Physiol. 248, 663-716. BLAKE MORE, C., VAN SLUYTER, R. C. & MovaHoN, J. A. (1976). Synaptic competition in the kitten's visual cortex. Cold Spring Harbor Symp. quaint. Biol. 40, The Synapse, pp. 601-609. BRODMANN, K. (1909). Vergleichende Lokalisationslehre der Gro88hirnrinde. Leipzig: J. A. Barth. BUISSEPRET, P. & IMBERT, M. (1976). Visual cortical cells: their developmental properties in normal and dark reared kittens. J. Physiol. 255, 511-525. CRAWFORD, M. L. J., BLAKE, R., COOL, S. J. & VON NOORDEN, G. K. (1975). Physiological consequences of unilateral and bilateral eye closure in macaque monkeys: some further observations. Brain Re8. 84, 150-154. Dow, B. M. (1974). Functional classes of cells and their laminar distribution in monkey visual cortex. J. Neurophysiol. 37, 927-946. Dow, B. M. & GOURAS, P. (1973). Color and spatial specificity of single units in rhesus monkey foveal striate cortex. J. Neurophysiol. 36, 79-100. DtRsTEILFR, M. R., GAREY, L. J. & MOVSHON, J. A. (1976). Reversal of the morphological effects of monocular deprivation in the kitten's lateral geniculate nucleus. J. Physiol. 261, 189-210. FINLAY, B. L., SCuLLER, P. H. & VOLMAN, S. F. (1976). Quantitative studies of single-cell properties in monkey striate cortex. IV. Corticotectal cells. J. Neurophysiol. 39, 1352-1361. GOURAS, P. (1974). Opponent-colour cells in different layers of foveal striate cortex. J. Physiol. 238, 583-602. GuILLERY, R. W. (1972). Binocular competition in the control of geniculate cell growth. J. comp. Neurol. 144, 117-129. HuIBEL, D. H. & WIEEL, T. N. (1962). Receptive fields, binocular interaction and functional architecture in the cat's visual cortex. J. Physiol. 160, 106-154. HUBEL, D. H. & WIESEL, T. N. (1963). Receptive fields of cells in striate cortex of very young, visually inexperienced kittens. J. Neurophysiol. 26, 994-1002. HUBEL, D. H. & WIESEL, T. N. (1968). Receptive fields and functional architecture of monkey striate cortex. J. Physiol. 195, 215-243. HUBEL, D. H. & WIESEL, T. N. (1970). The period of susceptibility to the physiological effects of unilateral eye closure in kittens. J. Physiol. 206, 419-436. HUBEL, D. H. & WIESEL, T. N. (1972). Laminar and columnar distribution of geniculo-cortical fibers in the macaque monkey. J. comp. Neurol. 146, 421-450. HUIBEL, D. H. & WIESEL, T. N. (1974). Sequence regularity and geometry of orientation columns in the monkey striate cortex. J. comp. Neurol. 158, 267-293. HUBEL, D. H. & WIESEL, T. N. (1977). Functional architecture of macaque monkey visual cortex. Proc. Roy. Soc. B 198, 1-59. HUBEL, D. H., WIESEL, T. N. & LEVAY, S. (1977). Plasticity of ocular dominance columns in monkey striate cortex. Phil. Tran8. R. Soc. Ser. B 278, 377-409. HUBEL, D. H., WIESEL, T. N. & STRYKER, M. P. (1977). Orientation columns in macaque monkey visual cortex demonstrated by the 2-deoxyglucose autoradiographic technique. Nature, Lond. 269, 328-330. JOHNSON, P. T., TUCEK, P. & MCGOWAN, M. A. (1977). Ocular fundus of the normal rhesus monkey. Vet. Med. Small Animal Clinician. 72, 645-648. LEVAY, S., HUBEL, D. H. & WIESEL, T. N. (1975). The pattern of ocular dominance columns in macaque visual cortex revealed by a reduced silver stain. J. comp. Neurol. 159, 559-576. LuND, J. S. (1973). Organization of neurons in the visual cortex, area 17, of the monkey (Macaca mulatta). J. comp. Neurol. 147, 455-495. MANSFIELD, R. J. WV. (1974). Neural basis of orientation perception in primate vision. Science N.Y. 186, 1133-1135. MARK, R. (1974). Memory and Nerve Cell Connections. Oxford: Clarendon Press. MERRILL, E. G. & AINsWORTH, A. (1 972). Glass-coated platinum-plated tungsten microelectrodes Med. & biol. Engng 10, 662-672. MICHAEL, C. R. (1972). Double opponent-color cells in the primate striate cortex. Physiologist, Wash. 15, 216.

REVERSAL OF MONOCULAR DEPRIVATION IN MONKEY

261

MITCHELL, D. E., CYNADER, M. & MOVSHON, J. A. (1977). Recovery from the effects of monocular deprivation in kittens. J. comp. Neurol. 176, 53-64. MOVSHON, J. A. (1975). The velocity tuning of single units in cat striate cortex. J. Physiol. 249, 445-468. MOVsHON, J. A. (1976). Reversal of the physiological effects of monocular deprivation in the kitten's visual cortex. J. Phyisol. 261, 125-174. MOVsHON, J. A. & BLmAEMORE, C. (1974). Functional reinnervation in kitten visual cortex. Nature, Lond. 251, 504-505. MOVSHON, J. A. & DtRsTELER, M. R. (1977). Effects of brief periods of unilateral eye closure on the kitten's visual system. J. Neurophysiol. 40, 1255-1265. NELsON, J. I., KAro, H. & BISHOP, P. 0. (1977). Discrimination of orientation and position disparities by binocularly activated neurons in cat striate cortex. J. Neurophysiol. 40, 260-283. OLsoN, C. R. & FREEMAN, R. D. (1975). Progressive changes in kitten striate cortex during monocular vision. J. Neurophysiol. 38, 26-32. PALMER, L. A. & ROSENQIST, A. C. (1974). Visual receptive fields of single striate cortical units projecting to the superior colliculus in the cat. Brain Res. 67, 27-42. PETPGREw, J. D. (1974). The effect of visual experience on the development of stimulus specificity by kitten cortical neurones. J. Physiol. 237, 49-74. POGGIO, G. F., BAKER, F. H., MANSFIELD, R. J. W., SILLITO, A. & GRIGG, P. (1975). Spatial and chromatic properties of neurons subserving foveal and parafoveal vision in rhesus monkey. Brain Res. 100, 25-59. RAxIc, P. (1977). Prenatal development of the visual system in rhesus monkey. Phil. Trans. R. Soc. Ser. B 278, 245-260. SCHILLER, P. H., FINLAY, B. L. & VOLMAN, S. F. (1976a). Quantitative studies of single-cell properties in monkey striate cortex. I. Spatiotemporal organization of receptive fields. J. Neurophysiol. 39, 1288-1319. ScHILLER, P. H., FnrAY, B. L. & VOLMAN, S. F. (1976b). Quantitative studies of single-cell properties in monkey striate cortex. 1I. Orientation specificity and ocular dominance. J. Neurophysiol. 39, 1320-1333. SHERK, H. & STRYKER, M. P. (1976). Quantitative study of cortical orientation selectivity in visually inexperienced kitten. J. Neurophysiol. 39, 63-70. SHERmAN, S. M., GuiLLERY, R. W., Kid, J. H. & SANDERsoN, K. J. (1974). Behavioral, electrophysiological and morphological studies of binocular competition in the development of the geniculo-cortical pathways of cats. J. comp. Neurol. 158, 1-18. SOKOLOFF, L. (1975). Influence of functional activity on local cerebral glucose utilization. In Brain Work. The Coupling of Function, Metabolismand Blood FlowintheBrain,ed. INGVAR, D. H. & LASSEN, N. A., pp. 385-388. Copenhagen: Munksgaard and New York: Academic Press. SPNzLLI, D. N., PRIBRAM, K. H. & BRIDGEMAN, B. (1970). Visual receptive field organization of single units in the visual cortex of the monkey. Int. J. Neurosci. 1, 67-74. STRYKER, M. & SHERK, H. (1975). Modification of cortical orientation selectivity in the cat by restricted visual experience: a reexamination. Science, N. Y. 190, 904-906. VITAL-DuIRAm, F., GAREY, L. J. & BLAKEMORE, C. (1978). Monocular and binocular deprivation in the monkey: morphological effects and reversibility. Brain Res. (In the Press). VON NOORDEN, G. K. (1973). Experimental amblyopia in monkeys. Further behavioral observations and clinical correlations. Invest. Ophthal. 12, 721-726. WIESEL, T. N. & HUBEL, D. H. (1963). Single-cell responses in striate cortex of kittens deprived of vision in one eye. J. Neurophysiol. 26, 1003-1017. WIESEL, T. N. & HuBEL, D. H. (1965a). Comparison of the effects of unilateral and bilateral eye closure on cortical unit responses in kittens. J. Neurophysiol. 28, 1029-1040. WIESEL, T. N. & HUBEL, D. H. (1965b). Extent of recovery from the effects of visual deprivation in kittens. J. Neurophysiol. 28, 1060-1072. WIEsL, T. N. & HUBEL, D. H. (1974). Ordered arrangement of orientation columns in monkeys lacking visual experience. J. comp. Neurol. 158, 307-318. WIESEL, T. N., HUBEL, D. H. & LAM, D. M. K. (1974). Autoradiographic demonstration of ocular-dominance columns in the monkey striate cortex by means of transneuronal transport. Brain Res. 79, 273-279. YATEs, J. T. (1974). Chromatic information processing in the foveal projection (area 8triata) of unanaesthetized primate. Vision Res. 14, 163-173.

262

C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND EXPLANATION OF PLATE

Nissl-stained section of a penetration site (LI) in V7710, showing the lamination scheme of Brodmann (1909) with a and fi subdivisions of layer W~c (Lund, 1973). WM = white matter. The oblique streak in the middle of the grey matter is part of the electrode track. Scale = 1 mm, corrected for shrinkage.

]Plate

The Journal of Physiology, Vol. 283 Z.

am,~*: 7N41Ki'>;

7-

. .I,. 11

III

IVa IVb

lVc

~-m

V

VI

WM

C. BLAKEMORE, L. J. GAREY AND F. VITAL-DURAND

(Facing p. 262)

The physiological effects of monocular deprivation and their reversal in the monkey's visual cortex.

J. Physiol. (1978), 283, pp. 223-262 With 1 plate and 17 text-figuree Printed in Great Britain 223 THE PHYSIOLOGICAL EFFECTS OF MONOCULAR DEPRIVATIO...
5MB Sizes 0 Downloads 0 Views