Receptive field characteristics of neurones in striate cortex of newborn lambs and adult sheep

NerrrosciefireVol. LO,No. 2, pp. 295-300, 1983 Printed in Great Britain

0306-4522/83 $3.00 + 0.00

Pergamon Press Ltd IBRO

RECEPTIVE FIELD CHARACTERISTICS OF NEURONES IN STRIATE CORTEX OF NEWBORN LAMBS AND ADULT SHEEP H. KENNEDY*, K. A. C. MARTINt and D. WHITTERIDCE~ Department of Experimental Psychology, South Parks Road, Oxford OX1 3UD, U.K. Abstract-The properties of cells of the striate visual cortex (Vl) have been studied in the normal adult sheep and in new-born lambs without visual experience. the majority of cells in the iamb are orientation specific, but 20”/, are non-oriented compared to only 3”/, cells in the adult. In the lamb there was little or no facilitation of binocularly-driven cells by simultaneous stimulation of both receptive fields. Cells which responded only to binocular stimulation of particular disparities (‘obligate binocular’ cells) were rarely encountered, In the adult, 15%of the sample were obhgate binocular cells and a further 28% showed binocular facilitation. Simple and complex receptive fields were found in similar proportions in both new-born lambs and adult sheep. End-stopped cells comprised 17”/:,of the sample in adults but only 2Td in the lambs. Direction sensitive cells were found more frequently in the sheep (21% of cells) than in the lamb (4% of cells). It is concluded that facilitatory processes in binocular cells and inhibitory mechanisms generally, seem much less developed in the lamb.

In this study we set out to examine the degree to which the properties of neurones in the striate cortex (visual area 1, or Vl) of the sheep are brought about by genetic factors as opposed to experience. When this problem was examined in the kitten, widely different results were reported. Hubel and Wiesel” and then Sherk and Stryker’9 found that although neurones were often sluggish and fatigued easily, their properties were similar to those of the mature cat. By contrast, other workers found that the majority of cells were not orientation-specific, that it was difficult to determine the receptive field (RF) boundaries and often impossible to show the organisation of the RF sub-regions.2,23 However, the kitten may not be the best choice of animal for this type of study because its visual cortex is so immature and the optic media are not clear at birth.‘,’ It would be more appropriate to choose an animal which is more mature at birth since in such an animal genetic processes would have had the greatest possible opportunity in which to act before the animal has visual experience. For this reason, we have chosen the sheep. Lambs have good visuomotor coordination on the day of birth, the optic media are clear, and almost all the neurones encountered respond briskly to visual stimuli.26 Thus, one can make a valid comparison of the RF characteristics in the new-born and adult animal EXPERIMENTAL PROCEDURES Acute experiments were carried out on 16 animals (7 lambs and 9 adults). The iambs were recorded within 12 h *Present address; INSERM-Unite 94. 16. avenue du Doven L&pine, 69500 Bron, France. tTo whom correspondence should be addressed. Abbreviations: RF, receptive field; Vi, visual area 1; V2, visual area 2; OB, obligate binocular. NSCIO/2

--i

295

of birth and were kept in the dark or with the eyes covered with opaque material prior to the experiment. The animals were anaesthetized with Ahhesin (Glaxo), or with halothane (Fluothane, ICI), and prepared for electrophysiological recording, when they were anaesthetized with Sagatal (May and Baker) and a 60:40 mixture of N,O and 0, and paralysed with an infusion of tubocurarine and gallamine triethiodide. Pupils were dilated with atropine and corneas were protected by contact lenses. The animals’ temperature was maintained at 3%39°C. Levels of anaesthesia after paralysis were adjusted so that separation of the hooves caused little change in blood pressure or heart rate.8b All electrode penetrations in adults and in the new-born were made in that part of area 17 situated in the ectolateral gyrus bounded by the lateral gyms medially and the ectolateral sulcus laterally. This area subserves vision from 4” lateral to the vertical meridian, then to 10” ipsilateral.” On the crown of the gyrus we sampled from the whole thickness of the cortex. In the medial wall of the ectolateral sulcus, needle tracks could run predominantly through either deep or superficial layers (Fig. 1). In both adults and new-born animals our sampling was biased towards the supragranular layers. This was not a disadvantage as these layers contain high proportions of obligate binocular ~ells.~.~~ End-tidal CO, was maintained at about 5%. Single units were isolated by means of glass-coated tungsten microelectrodes and were studied qualitatively with hand-held projectors and/or cutouts on a tangent screen placed at 114 cm in front of the animal. Comparison of quantitative and qualitative techniques have shown that for the RF features we have examined, (namely orientation tuning, RF dimensions, direction selectivity, orientation preference and end-zone inhibition) hand-plotting methods are adequate means of invest;gation.5~‘4~22~“~34 Onlv temnoral features such as velocitv iuning, and to a lesser exient directionality, require quantitative tests.*1,‘2We did not examine velocity tuning, and although hand-plotting may underestimate the numbers of cells having biases in their directional preference, the number of directional selective cells (which respond to movement in one direction only) can be accurately determined.** Dimensions of the RF were characterized by RF width. Receptive field width was measured from the minimum discharge field plotted with a narrow moving slit of light (0.2’) after optimisation of the stimulus parameters. Pre-

H. Kennedy

PI a/.

Fig. 1. Transverse section of the visual cortex in a new-born lamb. EC? ectolateral gyrus, ES: ectolateral sulcus, WM: white matter. Cortical layers numbered l-6 UC: unclassified, NO: non-oriented, UD: undriven, S & C defined in Experimental Procedures. Arrow at the bottom of the ectolateral sulcus marks the V,/V, border. One representative track with a lesion on the 5/6 border is shown. The unclassified cell was lost before it could be fully studied. ferred orientation and width of orientation tuning were hand-plotted with long narrow moving light or dark bars. except for end-stopped cells where shorter bars of optimal length were used. We cannot distinguish theoretically between axis and orientation selectivity in those cells which either were not tested with, or would not respond to, flashed stimuli. Nevertheless, the lack of direction preference along the preferred axis in most cells in the new-born lamb and the fact that many cells responded much more vigorously to long rather than short bars strongly suggest that these cells are orientation-specific rather than merely axis selective.*4 Two criteria were used to classify neurons: (1) the spatial disposition of subregions and (2) the effect of stimulus length on the strength of the response. The first criterion distinguished between RFs having overlapping and nonoverlapping subregions. A RF has overlapping subregions when the discharge regions of optimally oriented moving light and dark edges have a common spatial location and when a stationary flashed light slit reveals a single ‘ON-OFF’ region. A RF has non-overlapping subregions when moving dark and light edges and flashed stimuli reveal a single or separate ‘ON’ or ‘OFF’ area. Receptive fields with overlapping and non-overlapping subregions would

correspond to some extent to complex and simple cells in the classification scheme of Hubel and Wiesel” and will be referred to here as C and S cells. Analysis of the RF widths of each of these types did not allow A and B cells, as defined by Henry,” to be distinguished. The second criterion makes it possible to determine if a cell is endstopped. Cells showing a reduced response with increase of stimulus length were said to be partially endstopped and cells which showed a complete cessation of response with increasing stimulus length were said to be 100% endstopped. After carefully drawing a plot of the RF on the plane of the tangent screen the optimal stimulus was passed back and forth through the RF. If the response remained equal for both directions of movement, the neuron was classified as being non-directional. Neurons showing a marked increase in response level for one direction were classified as showing a directional preference (DP). Neurons which systematically showed an absence of response associated with a particular direction were classified as direction selective (DS). The ability of each eye to drive the cells was determined following alternate stimulation of each eye and classification of the cell in one of the seven ocular dominance groups.‘2 A Risley prism placed in front of the left

297

Visual cortex in new-born lambs and sheep eye, allowed the superimposition of the RFs in the right and left eye.*” Following simultaneous stimulation of the right and left eye, neurons were placed in one of four categories: (I) neurons showing no enhanced response to binocular stimulation, (2) neurons showing a weak enhancement of response, (3) neurons showing a strong ~han~ment and (4) neurons which could only be driven when both eyes were stimulated, such neurons were called obligate binocular (OB) cells. At the end of each experiment recording sites were marked with electrolytic lesions. Elecrode tracks and lesions were later identified in histological reconstructions. RESULTS

Virtually all the units encountered in the new-born lamb gave brisk and vigorous responses to visual stimulation. With flashed stimuli and moving edges it was possible to classify these cells as having overlapping and non-overlapping ‘ON’ and ‘OFF’ subregions and hence, except for those neurons which

did not show any orientation selectivity, it was possible to classify them as S or C cells. The proportion of S cells recorded in the adult and new-born animal were very similar (61 and 58x, respectively). Increase in stimulus length resulted in a decrease in response level in only 2”/, of the neurons recorded in these lambs. These neurons showed some residual activity when stimulated with long bars and were, therefore, classified as partially endstopped. The adult animal had many more endstopped units (16.7x), of which just under half showed 100% endozone inhibition. The ocular-dominance histograms obtained for new-born and adult animals were very similar (Fig. 2). The histograms include a group of cells in the adult that can only be driven if the RFs in both eyes are stimulated simultaneously.*” These 08 cells are rarely found in the newborn animal (Table 1). Further evidence was found of a difference in the interaction of ocular inputs in the cortex of young and mature animals. In the adult animal 28% of the cells showed an enhanced response if both eyes were stimulated simultaneously, where as only 6.1% of the cells encountered in the new-born animal showed such an effect. The mean width of the RF has been shown to be characteristic of S and C cells in the cat2’ and therefore a comparison of this aspect of RF dimension has been made for new-born and adult sheep. Receptive field widths were found to be narrower in the adult animal. The RF widths of S cells were narrower than that of C cells in both the new-born and mature animal. The RF widths of S cells in the

Fig. 2. Ocular dominance histogram for adult sheep (A) and new-born lambs (3). Ordinate: number of cells; abscissa: ocular dominance groups. Groups 1and 7 are driven solely by the contralateral and ipsilateral eye respectively. Group 4 cells are equally driven by both eyes and groups 2, 3 and 5, 6 are intermediate. OB: obligate binocular cells are those cells which can only be driven when both eyes are stimulated simu!taneously.

new-born animal were wider than in the adult (mean values 2.0” as opposed to 1.loo) as were C cells (mean values 2.5” as opposed to 1.68”). As shown in Fig. 3 the difference between the RF widths of S cells resulted from an absence in the new-born of RFs with a width less than 0.5”. As in the adult, the RFs of S cells in the new-born could contain more than one ‘ON’ or ‘OFF’ subfield. With an oblique penetration, a regular change of orientation tuning can be seen (Fig. 1) and the distribution of the range of orientation tuning in the mature and new-born animal is shown in Fig. 4. This figure includes all those cells which were reliably tested for orientation. Cells which failed to show orientation selectivity (non-oriented cells), are shown

“1fl

A

Fig. 3. Receptive field widths in (A) and (B) in the adult sheep and (C) and (D) in the new-born lamb. Ordinate: number of cells; abscissa: RFs width in degrees. (A) and (C) are RFs with non overlapping ‘ON’ and ‘OFF’ responses and (B) and (D) RFs with overlapping ‘ON-OFF’ responses. Vertical arrows refer to Mean values (A): 1.10”; (B): 1.68”; (C): 2.01”; (D): 2.55”.

Table 1. Response properties of cells recorded in the adult sheep and new-born lamb Obligate Binocular Cells New-born lambs Adult sheeo

Cells showing binocular facilitation Weak strong

Cells sensitive to direction of motion DP DS

End-stopped cells

Non-oriented cells

1.8 (1631 14.7

2.5 (163) 8.7

1.8 (163) 19.7

0.6 (163) II.9

3.6 (163) 8.7

1.8 (163) 16.7

(l?) 2.7

(218)

(218)

(218)

(126)

(126)

(126)

(150)

DP: directional preference; DS: direction selective. Figures in brackets refer to number of cells tested. Results given as percentages.

H. Kennedy

number of cells whose preferred orientation fell in one of 5 positions with respect to the principal axes. In the new-born animal, 387; of the cells had preferred orientations within 9” of a principal axis. The distribution of preferred orientations in the new-born animal was found to differ significantly from that of a random distribution (x2 test, P < 0.0005).

"Ii Fig. 4. Distribution of the widths of orientation tuning in (A) adult sheep and (B) in new-born lambs. Ordinate: percentage of total cells; abscissa: width of orientation tuning in degrees. N.0: cells which do not show orientation specificity. (A): 94 cells; (B): 144 cells.

in the last column. Clearly, the new-born animal differs from the adult in that 18.0% of its cells are non-oriented, whereas in the adult only 2.7 of the cells are non-oriented (Fig. 4). Although the adult animal has nearly three times as many cells as does the new-born lamb with an orientation tuning of less than 20’) a statistical comparison of the two distributions shows that the difference is only weakly significant (x* test, df 3, P < 0.025). However, the orientation range of the new-born and adult animal was found to be very different when S and C cells were considered separately (Table 2). Whereas, the mean orientation tuning of S cells in both sets of animals was very similar, that of C cells was considerably wider in the new-born animals. The mean orientation tuning of C cells in the adult was 40’ and was 55 in the new-born. Statistical analysis showed the differences of these two means to be significant (t-test, 0.05 P > 0.01). The preferred orientations of S and C cells in the adult were found to be evenly distributed. In the new-born animal however there was a marked tendency for the preferred orientation of the RFs to lie near to or on the horizontal or vertical axis. Preference for horizontal or vertical axis was as marked among S as among C cells. In Fig. 5 are shown the Table 2. Mean range of orientation tuning in degrees for S and C cells in adult sheep and new-born lamb

s cells c cells

Adult

New-born

N = 55 39.34 (2.60) N = 32 40.25 (3.27)

N = 55 39.56 (2.37) N = 36 55.02 (4.26)

Figures in brackets are the standard

et ul

tmms.

Fig. 5. Distribution of preferred orientations with respect to principal meridians. (A): adult sheep; (B): new-born lamb. Ordinate: number of cells; abscissa: number of degrees from the horizontal or vertical meridian.

DISCUSSION

The Frincipal conclusion that can be drawn from this study is that cortical neurones in the new-born lamb are considerably less specific in their stimulus requirements than those in the adult. In the lamb, there are very few directional or endstopped cells, some cells are less precisely tuned to orientation, RFs tend to be larger and there is little facilitation when both eyes are stimulated simultaneously. In addition, there is an over-representation in the lamb of cells tuned to orientations near or on a principal axis. It is clear from our results, however, that some basic features of the receptive field are present at birth. Cells in the new-born animal can be classified as S and C and are found in similar proportion as in the adult, the majority are orientation-specific and their cumulative ocular dominance histogram is similar to that of the adult. The brisk responses of cells recorded in the lamb contrast with the sluggish, poorly driven and easily fatigued cells commonly found in kittens 2.h.7.13.23and suggest a greater maturity of cortical connections in the lamb. Three aspects of the characteristics are of particular interest in relation to the cat. Receptive .field type Although Hubel and Wiesel mention recording from both simple and complex cells in the visual cortex of the young kitten,” there are two reports of an almost total absence of complex cells in very young kittens.3.4 We found similar proportions of S and C cells in Vl subserving central vision in the adult sheep as in the adult cat.*” The proportion of S and C cells were identical in the new-born and adult sheep. Orientation

selectivity

Estimates of the orientation selectivity of cel!s in the young kitten range from the claim of “very few orientation selective units”‘,23 to the finding that “preference in stimulus orientation was common to all of the units isolated.“‘3 Others, using quantitative and qualitative methods have obtained estimates falling between these two extremes.4,6,7,29 Many of the discrepancies may be due to differences in criteria. One new-born monkey has also been recorded from and all 23 cells were found to be orientation-specific.” In the lamb, just over 80% of the units encountered were orientation specific compared to about 967; in the adult. Interestingly, a higher proportion of cells (95”/,) recorded in V2 of the newborn lamb are

Visual cortex in new-born lambs and sheep

299

second visual area in the lamb also shows this orientation selective.*b This could occur if the develtendency.8b Thus, although RF disparities are opment of V2 begins earlier than that of Vl. The present, the absence of the facilitation, which is usual orientation tuning of S cells appears to be entirely established by innate process and these cells show in the adult, in the cell’s response to binocular stimulation at different distances may indicate that similar orientation tuning ranges in the new-born lamb and adult sheep. C cells, however, have the lamb has poor stereopsis. The new-born lamb can significantly wider orientation tuning ranges in the however, avoid a visual cliff,32 but stereopsis may not new-born animal which may make them more sus- be necessary for this form of depth discrimination. The tuning of cells to disparity is probably brought ceptible to environmental influences4.” The adult sheep showed an even distribution of about by structural changes involving both excitatory and inhibitory connections. This process may well orientation preferences. This contrasts with the adult cat where neurons situated in cortex subserving cen- require visual experience to compensate for the large tral vision have more orientation preferences for changes in eye position which occur during devefstimuli along the vertical and horizontal axisi5*‘9.24.28opment in the sheep. Disparity tuning apart, those characteristics of the and this may be more marked in the kitten. The receptive field which were found to be less specific in new-born lamb differed from the adult animal in that both S and C cells showed a marked tendency for the lamb, have been shown to be dependent on orientation preferences to be clustered around the intra-cortical inhibition, which is probably mediated acid (GABA). ho~zontal and vertical axes. Our results would sup- by the transmitter, ~-aminobutyric Application of GABA antagonists to single cells gives port the suggestion that orientational anisotropy rise to a disappearance of directionality, end-stopping when present in adults may be an innate feature and not the result of over-exposure to horizontal and and a broadening or suppression of orientation tunThe appearance of more selective RF charvertical stimuli.16 As Fregnac and Imbert have sug- ing. 27,30,3’ gested, these initial biases may reflect the manner in acteristics in the adult may thus be due to the which orientation preferences are built up in the development of the inhibitory connections in the cortex. Such an interpretation would be consistent cortex during development.” with structural data in the kitten where symmetric Disparity tuning synapses (thought to be inhibitory) initially are virtually absent.35 However, if this is the reason then our Only one study has addressed the issue of disparity tuning in the kitten. Pettigrew was unable to deter- data requires that the inhibitory mechanisms for mine RF disparity because the borders of the RFs generating orientation specificity would have to mawere diffuse and the RF of the non-dominant eye ture far earlier than other inhibitory processes. could often not be defined at al1.23Although there The slightly larger RFs found in the lamb may be was strong binocular facilitation in the kitten, this due in part to growth of the eye which in the cat has facilitation occurred over a very wide range of dis- been shown to reduce the size of the retinal ganglion parity and in no sense could the cells be said to be cell RFs.**” Changes in the size of the cortical arbortuned to binocular disparity. Eye movements could isations of the cells of the lateral geniculate nucleus also not be easly monitored because the cloudy optics may also reduce the RF size.” of the kitten make it impossible to plot any landmarks. By contrast, in the lambs virtually no sumacknowledgements-me wish to thank MS A. Wimshurst mation, facilitation or inhibition was found on binocfor technical assistance. The work was supported by the ular stimulation, but field borders could be accurately M.R.C. H. Kennedy was partially supported by the Wellplotted, and in some cases, showed disparities. The come Trust. REFERENCES 1. Anker R. L. and Cragg B. G. (1974) Development of the extrinsic connections of the visual cortex in the cat. J. camp. Neural. 154,2942. 2. Barlow H. B. and Pettigrew J. D. (1971) Lack of specificity of neurons in the visual cortex of young kittens. J. P&&i., Lond. 218, 98P. 3. Beckmann R. and Albus K. (1982) The geniculocortical system in the early postnatal kitten: an electrophysiological investigation. Exp. Bruin Res. 47, 49-56. 4. Blakemore C. and Van Sluyters R. C. (1975) Innate and environmental factors in the development of the kitten’s visual cortex. J. Physiol., Lond. 248, 663-716. 5. Blasdel G. G., Mitchell D. E., Muir D. W. and Pettigrew, J. D. (1977) A physiological and behavioural study in cats of the effects of early visual experience with contours of a single orientation. J. Physiof., Land. 265, 615-636. 6. Bonds A. B. (1979) Developm~t of orientation tuning in the visual cortex of kittens. In ~ff~elffp~enf~i ~~urobjo~Qgy of Vision (ed. Freeman R. D.) NATO A.S.I.S. (series A. Life Sci. 27, 31-41). 7. Buisseret P. and Imbert M. (1976) Visual cortex cells: their developmental properties in normal and dark reared kittens. J. Physiol., Land. 255, 511-525. 8a. Clarke, P. G. H., Donaldson, I. M. L. and Whitteridge, D. (1976) Binocular visual mechanisms in cortical areas I and II of the sheep. .I. Physiol., Land. 256, 509-526. 8b. Clarke P. G. H., Ramachandran V. S. and Whitteridge D. (1979) The development of the binocular depth cells in the secondary visual cortex of the lamb. Proc. R. Sot. Land. B. 204, 455465.

300

H. Kennedy

er (II.

8c. Clarke. P. G. H. and Whitteridae. D. (1976) The cortical visual areas of the sheen J. Phvsiol.. Land. 256. 497-508. 9 Cragg B. G. (1975) The development of synapses in the visual system of the cat.‘.!. co&. N&01. 160, 147-166. 10 Fregnac Y. and Imbert M. (1978) Early development of visual cortical cells in normal and dark reared kittens: relationship between orientation selectivity and ocular dominance. J. Physiol., Lond. 278, 27-44. 11, Henry G. H. (1977) Receptive field classes of cells in the striate cortex of the cat. Bruin Res. 133, L-28. and functional architecture in the cat’s 12. Hubel D. H. and Wiesel T. N. (1962) Receptive fields. binocular interactjon visual system. _I. Physiol., Land. 160, 106154. 13. Hubel D. H. and Wiesel T. N. (1963) Receptive fields of cells in striate cortex of very young, visually inexperienced kittens. J. Neurophysiol. 26, 994-1002. and simple/complex cell classifications in cat striate 14. Kato H. and Bishop P. O., Orban, G. A. (1978) Hypercomplex cortex. J. Neurophysiol. 41, 1071-1096. in cat visual cortical neurons. 15. Kennedy H. and Orban G. A. (1979) Preferences for horizontal or vertical orientation .I. Physiol., Lond. 296, 61P. A., Brill S. and Held R. (1975) O~entation anisotropy infant vision. Science 191), 16. Leehey S. C., Moskowitz-Cook 900-90 t in layer IV of the 17. LeVay S., Stryker M. P. and Shatz C. J. (1978) Ocular dominance columns and their development cat’s visual cortex: a quantitative study. J. camp. Neural. 179, 223-244. sensitivity and binocularity 18. Leventhal A. G. and Hirsch H. V. B. (1977) Effects of early experience upon orientation of neurons in visual cortex of cats. Proc. natn. Acad. Sci. U.S.A. 74, 1272-1276. 19. Leventhal A. G. and Hirsch H. V. B. (1980) Receptive field properties of different classes of neurons in the visual cortex of normal and dark-reared cats. J. ~euruphy~~~ol. 43, 1I1 1-l 132. selectivity 20. Orban G. A. and Kennedy H. (1981a) The influence of eccentricity on receptive field types and orientation in areas 17 and 18 of the cat. Brain Res. 208, 203-208. 21. Orban G. A., Kennedy H. and Maes H. (198 I b) Response to movement of neurons in areas 17 and I8 of the cat: velocity sensitivity. J. Neurophysiol. 45, 1043-1058. 22. Orban G. A., Kennedy H. and Maes H. (1981~) Response to movement of neurons in areas 17 and 18 of the cat: direction selectivity. 1. Neurophysiol. 45, 1059-1073. of stimulus specificity by kitten cortical 23. Pettigrew J. D. (1974) The effect of visual experience on the development neurons. J. Physiol., Lond. 237, 49-74. 24. Pettigrew J. D., Nikara T. and Bishop P. 0. (1968) Responses to moving slits by single units in cdt striate cortex. E.Y/I. Brain Res. 6, 313-390. V. S., Clarke P. G. H. and Whitteridge D. (1977) Cells selective to binocular disparity in the cortex 26. Ramachandran of newborn lambs. Nature, New Biol. 268, 333-335. 27. Rose D. and Blakemore C. (1974) Functions of inhibition in visual cortex: effects of bicuculline. Nature. Lond. 249, 375-377. selectivity in the cat’s visual cortex. E.x~. Brain Rex 20, 28. Rose D. and Blakemore C. (1975) An analysis of orientation lO99-1113. 28a. Rusoff, A. C. and Dubin, M. W. (1977) Development of receptive-held properties of retinal ganglion cells in kittens. J. Neurophysiol. 37, I 188-l 198. 29. Sherk H. and Stryker M. P. (1976) Quantitative study cortical orientation selectivity in visually inexperienced kitten. J. Neurophysiol. 39, 63-70. 30. Sillito A. M. (1975) The effectiveness of bicuculline as an antagonist of GABA and visually evoked inhibition in the cat’s striate cortex. J. Phyxioi.. Land. 250, 287-307. 31. Sillito A. M., Kemp J. A., Milson J. A. and Berardi N. (1980) A re-evaluation of the mechanisms underlying simple cell orientation selectivity. Brain Res. 194, 517-520. 32. Walk R. D. and Gibson E. J. (1961) A comparative and analytical study of visual depth perception, Psychol. Mono~r. 75(15). 144. 33. Wiesel T. N. and Hubel D. H. (1974) Ordered arrangements of orientation columns in monkeys lacking visual experience. J. romp. Neural. 158, 307-318. 34. Wilson J. R. and Sherman S. M. (1976) Receptive field characteristics of neurons in cat striate cortex: changes with visual field eccentricity. J. ~~e~rophy.sio~. 39, 5-12-533. 35. Winfield D. A. (1981) The postnatal development of synapses in the visual cortex of the cat and the effect of eyelid closure. Brain Res. 206, 166171.

of

(Accepted

16 Mq

1983)