Evidence for separate pathways within the tecto-geniculate projection in the tree shrew.

Two layers (3 and 6) in the dorsal lateral geniculate nucleus (GLd) of the tree shrew (Tupaia belangeri) receive projections from the superficial layers of the superior colliculus. The goal of this study was to determine whether the same or different cells in the superior colliculus give rise to the projections to layers 3 and 6 by following individual axons labeled with biocytin from the superior colliculus to the GLd. The results show that the terminal fields differ in the two layers--those in layer 3 are restricted to a line of projection, whereas those in layer 6 are elongated along the dimension orthogonal to a line of projection. Another important difference between axons that project to GLd layers 3 and 6 is that those that project to layer 6 give off collaterals to the posterior pretectal nucleus, whereas at least some axons that project to layer 3 send a collateral to the ventral lateral geniculate nucleus (GLv). These results suggest that the superior colliculus exerts separate influences on these two GLd layers, both of which project to separate targets above layer IV in the striate cortex. The biocytin method has proved useful by showing the dendritic trees of the superior colliculus cells of origin, the pathways taken by the axons (including the presence of collaterals), and the terminal fields both within and outside the GLd.

Terminal arbors of individual, physiologically identified geniculocortical axons in the tree shrew's striate cortex.

We studied the terminal patterns of single, physiologically identified geniculocortical axons in the striate cortex of the tree shrew by using intracellular recording and labeling methods. Axons were classified by their response to the onset (ON-center) or offset (OFF-center) of a light stimulus presented to the ipsilateral or contralateral eye. Then, we attempted to penetrate each axon for labeling with horseradish peroxidase. We recovered 23 axons and studied 16 of these in detail. Light microscopic reconstructions of these axons revealed several distinct terminal patterns within cortical layer IV. ON-center axons had terminal arbors that ended mainly in the upper part of layer IV (IVa), while OFF-center axons ended in the lower part of layer IV (IVb). Within layers IVa and IVb, axons driven by the ipsilateral eye and those driven by the contralateral eye had overlapping distributions. However, their terminal arbors differed in size, in shape, and in the number of boutons. Compared with contralateral eye arbors, ipsilateral eye axons were on average three times larger in lateral extent (925 microns vs. 325 microns), spread over four times the surface area (0.13 mm2 vs. 0.03 mm2), and supported one and one-half times as many terminal boutons (1,647 vs. 1,086). The ipsilateral eye axons had more boutons at the edges of layer IV (i.e., the upper part of layer IVa and the lower part of layer IVb), while those from the contralateral eye axons were more evenly distributed. These results show that each functional class of geniculocortical fiber has a different laminar and areal arrangement of boutons and we consider the significance of these differences for visual cortical function.

Sublaminar organization within layer VI of the striate cortex in Galago.

In this study we examined the organization of projections from the striate cortex to the dorsal lateral geniculate (GL) and pulvinar (PUL) nuclei in the prosimian Galago by using retrograde transport methods. Injections of wheat germ agglutinin-conjugated horseradish peroxidase (WGA-HRP) into the PUL labeled two bands of cells in the striate cortex: the first consisted of large pyramidal cells in the upper half of layer V; the second consisted of small and medium-size pyramidal cells located in the deepest part of layer VI. The location of cells within layer VI coincided with a clear cytoarchitectonic sublayer, VIb, which contains fewer and paler staining cells than VIa. Injections of WGA-HRP involving all layers of the GL produced an uninterrupted band of pyramidal cells distributed throughout layer VI (a and b), including the region labeled after injections into the PUL. Thus as a first approximation, layer VI can be divided into an upper tier (VIa) that projects only to the GL and a lower tier (VIb) that projects to both the GL and PUL. Injections of WGA-HRP that were restricted to one or a few GL layers revealed a further refinement of the subdivisions within layer VI. Injections into the parvicellular and intercalated (or koniocellular) layers of the GL labeled neurons predominantly in the upper half of layer VIa, whereas injections restricted to the magnocellular layers labeled neurons in the lower half of layer VIa and in layer VIb. In order to determine whether individual neurons in layer VIb send axon collaterals to both the GL and PUL, we injected WGA-HRP into one nucleus and fluorescent rhodamine latex beads into the other. In three experiments, we found only one double-labeled cell. In sum, the results provide evidence that layer VI is divided into at least three sublayers: upper VIa, which projects to the intercalated and parvicellular GL layers; lower VIa, which projects to the magnocellular GL layers; and VIb, which sends separate projections to the magnocellular layers of the GL and to the PUL. The segregation observed is sufficiently discrete to propose the existence of multiple, descending pathways from layer VI of the striate cortex that complement those ascending from the GL and PUL.

Innervation patterns of single physiologically identified geniculocortical axons in the striate cortex of the tree shrew.

We examined the termination patterns of single geniculocortical axons in the striate cortex of the tree shrew by using intracellular recording and horseradish peroxidase staining methods. Axons were classified by whether they responded to light onset (ON center) or light offset (OFF center) and whether they were driven by the ipsi- or contralateral eye. Afferents with ON-center responses end in the upper part of layer IV (IVa) whereas afferents with OFF-center responses end in the lower part of layer IV (IVb). Within each tier, axons driven by the ipsilateral and contralateral eye overlap. These results suggest that binocular convergence occurs within layer IV without mixing the information from the ON- and OFF-center pathways and we consider the significance of this arrangement for visual cortical function.

Cholinergic and monoaminergic innervation of the cat's thalamus: comparison of the lateral geniculate nucleus with other principal sensory nuclei.

The cholinergic and monoaminergic innervation of the lateral geniculate nucleus (GL) and other thalamic nuclei in the cat was examined by using immunocytochemical and tract-tracing techniques. Cholinergic fibers, identified with an antibody to choline acetyltransferase (ChAT), are present in all layers of the GL. They are fine in caliber and exhibit numerous swellings along their lengths. The A layers, the magnocellular C layer, and the medial interlaminar nucleus are rich in cholinergic fibers that give rise to prominent clusters of boutons, while the parvicellular C layers contain fewer fibers that are more uniformly distributed. The interlaminar zones are largely devoid of ChAT-immunoreactive fibers. Double-label experiments show that cholinergic projections to the GL originate from two sources, the pedunculopontine reticular formation (PPT) and the parabigeminal nucleus (Pbg). The PPT contributes cholinergic fibers to all layers, while Pbg projections are limited to the parvicellular C layers. The lateral geniculate nucleus has a much greater density of cholinergic fibers than the other principal sensory nuclei: the density of fibers in the A layers is more than three times greater than that in the ventral posterior nucleus (VP) or the ventral division of the medial geniculate nucleus (GMv). In contrast, serotonin (5-HT)-immunoreactive fibers are distributed with equal density across the principal thalamic nuclei, while tyrosine hydroxylase (TH)-immunoreactive fibers (presumed to contain norepinephrine) are noticeably less dense in the GL than in the others. Monoaminergic fibers also differ from cholinergic fibers in their laminar distribution within the GL: both TH- and 5HT-immunoreactive fibers are distributed evenly across the layers and interlaminar zones and are slightly more abundant in the parvicellular C layers than in the other layers. Other thalamic nuclei rich in cholinergic fibers include the pulvinar nucleus, the ventral lateral geniculate nucleus, the intermediate nucleus of the lateral group, the lateral medial and suprageniculate nuclei (Graybiel and Berson: Neuroscience 5:1175-1238, '80), and the paracentral and central-lateral components of the intralaminar nuclei. This pattern matches the distribution of projections from the PPT and is similar, but not identical, to the pattern of acetylcholinesterase staining. The fact that most of the nuclei rich in cholinergic fibers have been implicated in visual sensory or visual motor functions suggests that cholinergic projections from the reticular formation play an especially important role in visually guided behavior.

Organization of cholinergic synapses in the cat's dorsal lateral geniculate and perigeniculate nuclei.

In the preceding article, we showed that cholinergic fibers originating from the brainstem reticular formation provide a dense innervation of the lateral geniculate nucleus. In this report we describe the ultrastructure of these fibers and their relations with other elements in the neuropil of the lateral geniculate nucleus. Cholinergic fibers were labeled with an antibody to choline acetyltransferase (ChAT), the synthesizing enzyme for acetylcholine (ACh). In the A-laminae of the lateral geniculate nucleus, ChAT + profiles are small and contain tightly packed, mostly round vesicles. Some end in encapsulated synaptic zones where they form asymmetrical synaptic contacts with processes of both projection cells and interneurons. Others form synapses upon the shafts of dendrites. Of the four classical types of vesicle-containing profiles identified by Guillery (Z. Zellforsch. Mikrosk. 96:1-38, '69; Vision Res. [Suppl.] 3:211-227, '71), ChAT + profiles most closely resemble RSD profiles (Round vesicles, Small profile, Dark mitochondria). However, as a population, ChAT + profiles can be distinguished from the unlabeled population of RSD profiles because they are larger in size, contain more mitochondria, and make synapses with smaller postsynaptic membrane specializations. Each of these differences is statistically significant and together they indicate that ChAT + profiles are a distinct morphological type of synaptic profile. ChAT + profiles in the perigeniculate nucleus resemble those found in the lateral geniculate nucleus; they also make synapses with obvious postsynaptic thickenings.

Cholinergic innervation of the superior colliculus in the cat.

The superficial and intermediate gray layers of the superior colliculus are heavily innervated by fibers that utilize the neurotransmitter acetylcholine. The distribution, ultrastructure, and sources of the cholinergic innervation of these layers have been examined in the cat by using a combination of immuno-cytochemical and axonal transport methods. Putative cholinergic fibers and cells were localized by means of a monoclonal antibody to choline acetyltransferase (ChAT). ChAT immunoreactive fibers are distributed throughout the depth of the superior colliculus, with particularly dense zones of innervation in the upper part of the superficial grey layer and in the intermediate grey layer. Within the superficial grey layer, the fibers form a continuous, dense band, whereas within the intermediate grey layer the fibers are arranged in clusters or patches. Although the patches are present throughout the rostrocaudal extent of the superior colliculus, they are most prominent in middle to caudal sections. The structure of the ChAT immunoreactive terminals was examined electron microscopically. The appearance of the terminals is similar in the superficial and intermediate grey layers. They contain closely packed, mostly round vesicles, and form contacts with medium-sized dendrites that exhibit small, but prominent postsynaptic densities; a few of the terminals contact vesicle-containing profiles. To identify the sources of the cholinergic input to the superior colliculus, injections of wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP) were made in the superior colliculus and the sections were processed to demonstrate both the retrograde transport of WGA-HRP and ChAT immunoreactivity. Neurons containing both labels were found in the parabigeminal nucleus, and in the lateral dorsal and pedunculopontine tegmental nuclei of the pontomesencephalic reticular formation. Almost every cell in these nuclei that contained retrograde label was also immunoreactive for ChAT. The similarities between the laminar distributions of the ChAT terminals and the terminations of the pathway from the parabigeminal nucleus (Graybiel: Brain Res. 145:365-374, '78) support the view that the latter nucleus is a source of the cholinergic fibers in the superficial grey layer. The possibility that the pedunculopontine tegmental nucleus is a source of cholinergic fibers in the deep layers was tested by examining the distribution of labeled fibers following injections of WGA-HRP into this region of the tegmentum. Patches of labeled terminals were found in the intermediate grey layer that resemble in distribution the patches of ChAT immunoreactive fibers in this layer.(ABSTRACT TRUNCATED AT 400 WORDS)

Morphology of retinogeniculate X and Y axon arbors in cats raised with binocular lid suture.

1. We examined the terminal arbors of single, physiologically identified retinogeniculate X and Y axons in 13 adult cats raised from birth with binocular lid suture. We recorded in the optic tract from 146 retinogeniculate axons. We studied the response properties of each axon encountered and attempted to penetrate it for labeling with horseradish peroxidase. 2. We attempted to classify each retinogeniculate axon as X or Y on a standard battery of tests. We thus identified 46 X and 91 Y axons; 5 axons had unusual response properties, and 4 axons were lost before they could be adequately identified. The X and Y axons had response properties that were completely normal by our criteria. The 5 unusual axons exhibited linear spatial and temporal summation, which is a property of X cells, despite all of their other tested response properties being consistent with those of Y cells. 3. We achieved complete, dark labeling of 13 X and 13 Y axons that form the data base for all of our qualitative and quantitative morphological observations. All of these labeled axons had response properties entirely normal for their X or Y class. Nine of the labeled X axons arise from the contralateral retina and 4 from the ipsilateral retina, whereas the respective numbers for the Y axons are 8 and 5. 4. Each of the individual retinogeniculate X axons form terminal arbors that appeared essentially normal in terms of location within geniculate lamina A or A1, shape, volume, and number of terminal boutons. 5. In contrast, the retinogeniculate Y axons form clearly abnormal arbors with diminished projections, both in terms of bouton numbers and arbor volumes. For Y axons from the contralateral retina, a roughly normal arbor is formed in the C-laminae, despite greatly diminished or absent projections formed in lamina A, something never seen in normal cats. For Y axons from the ipsilateral retina, the projections to lamina A1 are also diminished, and the arbors there are all limited to the ventral half of the lamina, a pattern rarely seen for normal Y axons. 6. The selective reduction in retinogeniculate Y axon arbors in these binocularly lid-sutured cats is consistent with similar observations reported for monocularly lid-sutured and strabismic cats but is quite different from the apparently normal development of retinogeniculate axon arbors in cats raised in complete darkness.(ABSTRACT TRUNCATED AT 400 WORDS)

Synaptic circuitry of physiologically identified W-cells in the cat's dorsal lateral geniculate nucleus.

The cat's retinogeniculocortical system is comprised of at least 3 parallel pathways, the W-, X-, and Y-cell pathways. Prior studies, particularly at the level of the lateral geniculate nucleus, have focused on X- and Y-cells. In the present study, we describe the synaptic inputs for 2 geniculate W-cells from the parvocellular C-laminae after these neurons were physiologically identified and intracellularly labeled with HRP. For each of the W-cells, we examined electron micrographs taken from over 500 consecutive thin sections; we reconstructed the entire soma plus roughly 15% of the dendritic arbor and determined the pattern of synaptic inputs to these reconstructed regions of each neuron. In several ways, each W-cell exhibits a similar pattern of synaptic inputs. First, we estimate that each W-cell receives approximately 3000-4000 synaptic contacts, which occur most densely on dendrites 50-150 microns from each soma. Second, axosomatic contacts are extremely rare, and most derive from terminals with flattened or pleomorphic vesicles (F terminals). Third, terminals with round vesicles, large profiles, and pale mitochondria (RLP terminals), which are presumed to be retinal terminals, form only about 2-4% of all synapses onto these W-cells; these synapses occur on proximal dendrites. Fourth, F terminals, which provide roughly 15-20% of all synaptic input to these cells, occupy the same region of proximal dendritic arbor as do the RLP terminals. Fifth, and finally, terminals with round vesicles, small profiles, and dark mitochondria (RSD terminals) provide the majority of synapses along all portions of the dendritic arbor. Compared with geniculate X- and Y-cells of the A-laminae (Wilson et al., 1984), these W-cells are innervated by fewer synapses overall and, in particular, by dramatically fewer synapses from RLP (or retinal) terminals. This paucity of direct retinal input to geniculate W-cells might explain the remarkably poor responsiveness of these neurons to visual stimuli and to electrical activation of the optic chiasm.

Synaptic circuits involving an individual retinogeniculate axon in the cat.

In order to describe the circuitry of a single retinal X-cell axon in the lateral geniculate nucleus, we physiologically characterized such an axon in the optic tract and injected it intra-axonally with horseradish peroxidase. Subsequently, we recovered the axon and employed electron microscopic techniques to examine the distribution of synapses from 18% of its labeled terminals by reconstructing the unlabeled postsynaptic neurons through a series of 1,200 consecutive thin sections. We found remarkable selectivity for the axon's output, since only four of the 43 available neurons in a limited portion of the terminal arbor receive synapses from labeled terminals. Moreover, the distribution of these synapses on the four neurons, which we term cells 1 through 4, varies with respect to synapses from other classes of terminals that contact the same cells, including synapses from unlabeled retinal terminals. For cells 1 and 3, the labeled terminals provide 49% and 33%, respectively, of their retinal synapses, and these are located on both dendritic shafts and appendages. Synapses from the injected axon to these cells are thus integrated with those from other retinal axons. For cell 2, the labeled terminals provide 100% of its retinal synapses, but these synapses converge on clusters of dendritic appendages where they are integrated with convergent inhibitory inputs. Finally, for cell 4, the labeled terminals provide less than 2% of its retinal inputs, and these are distally located; we suggest that these synapses are remnants of physiologically inappropriate miswiring that occurs during development. The findings from this study support a concept of selectivity in neuronal circuitry in the mammalian central nervous system and also reveal some of the diverse integrative properties of neurons in the lateral geniculate nucleus.

Morphology and physiology of single neurons in the medial interlaminar nucleus of the cat's lateral geniculate nucleus.

Physiological studies have shown that the cat's retinogeniculocortical system is comprised of at least three parallel and independent pathways, the W-, X-, and Y-cell pathways. The morphological correlates of the constituent W-, X-, and Y-cells have been determined both in the retina and in the A and C laminae of the lateral geniculate nucleus. The aim of this study was to extend these structure/function relationships to neurons in laminae 1 and 2 of the medial interlaminar nucleus (MIN), which is a division of the cat's dorsal lateral geniculate nucleus. We used intracellular injection of horseradish peroxidase (HRP) into individual, physiologically identified MIN neurons. Since this procedure may yield an unrepresentative sample of MIN neurons, two controls were performed. First Nissl staining showed that the soma sizes of intracellularly labeled cells were representative of those of all MIN cells. Second, retrograde labeling following HRP injections into the optic radiations or specific visual cortical areas showed that the intracellularly labeled MIN cells were representative of MIN relay neurons. Many of the retrogradely labeled cells were so well filled that their entire dendritic arbors were revealed. Of 70 MIN neurons recorded physiologically, 22 were injected with HRP and successfully recovered. We also completely labeled the somata and dendrites of 114 MIN neurons from HRP injections into the optic radiations and retrogradely labeled 165 MIN neurons by injection of HRP into visual cortical areas. Our sample of intracellularly injected neurons, which were all Y-cells, were morphologically representative of all MIN relay cells. We thus conclude that laminae 1 and 2 of the MIN contain a nearly homogeneous population of Y-cells with properties essentially identical to those of Y-cells in the A and C laminae of the lateral geniculate nucleus. When viewed in the coronal plane, MIN projection neurons typically exhibited oval or elongated somata. In the medial and ventral parts of the MIN, these somata were smaller and more flattened. MIN soma sizes extended over the full range of those seen in the A laminae. Dendritic arbors of most MIN relay neurons radiated in a fairly spherical fashion. In the medial and ventral parts of the MIN, however, dendrites were oriented in a more bipolar fashion, but intermediate forms between spherical and bipolar arbors were also common. Dendrites of MIN projection neurons were typically smooth; most primary dendrites were straight, but secondary dendrites were more variable in structure.(ABSTRACT TRUNCATED AT 400 WORDS)

Synaptic connectivity of a local circuit neurone in lateral geniculate nucleus of the cat.

Although receptive fields of relay cells in the lateral geniculate nucleus of the cat nearly match those of their retinal afferents, only 10-20% of the synapses on these cells derive from the retina and are excitatory. Many more (30-40%) are inhibitory and largely control the gating of retinogeniculate transmission. These inhibitory synapses derive chiefly from two cell types: intrinsic local circuit neurones and cells in the adjacent perigeniculate nucleus. It has been difficult to study the functional organization of these inhibitory pathways; most efforts have relied on indirect approaches. Here we describe the use of direct techniques to study a local circuit neurone by iontophoresing horseradish peroxidase (HRP) into it, which completely labels the soma and processes of cells for subsequent light- and electron microscopic analysis. Although the response properties of the labelled cell are virtually indistinguishable from those of many relay cells, its morphology is typical of 'class 3' neurones (see Fig. 1 legend), which are widely believed to be interneurones (but see ref. 12). Here, we refer to the cell as a 'local circuit neurone', which allows for the possibility of a projection axon, rather than as an 'interneurone', a term that commonly excludes a projection axon. We find that the labelled cell has a myelinated axon, but that the axon loses its myelin within 50 microns of the soma and has not yet been traced further. The dendrites of the labelled cell possess presynaptic terminals that act as intrinsic sources of inhibition on geniculate relay cells. We also characterize other morphological aspects of this inhibitory circuitry.

The projections of single thalamic neurons onto multiple visual cortical areas in the cat.

Fluorescent dyes Fast Blue and Nuclear Yellow injected into pairs of visual cortical and parietal 'association' cortical areas in the cat revealed the presence of retrogradely double-labeled cells in the intralaminar nuclei and lateral posterior-pulvinar complex of the thalamus. These results demonstrate the projection of individual thalamic neurons onto multiple cortical areas.

Connections of the multiple visual cortical areas with the lateral posterior-pulvinar complex and adjacent thalamic nuclei in the cat.

The present report describes the patterns of cat thalamocortical interconnections for each of the 13 retinotopically ordered visual areas and additional visual areas for which no retinotopy has yet emerged. Small injections (75 nl) of a mixture of horseradish peroxidase and [3H]leucine were made through a recording pipette at cortical injection sites identified by retinotopic mapping. The patterns of thalamic label show that the lateral posterior-pulvinar complex of the cat is divided into three distinct functional zones, each of which contains a representation of the visual hemifield and shows unique afferent and efferent connectivity patterns. The pulvinar nucleus projects to areas 19, 20a, 20b, 21a, 21b, 5, 7, the splenial visual area, and the cingulate gyrus. The lateral division of the lateral posterior nucleus projects to areas 17, 18, 19, 20a, 20b, 21a, 21b, and the anterior medial (AMLS), posterior medial (PMLS), and ventral (VLS) lateral suprasylvian areas. The medial division of the lateral posterior nucleus projects to areas AMLS, PMLS, VLS, and the anterior lateral (ALLS), posterior lateral (PLLS), dorsal (DLS) lateral suprasylvian areas, and the posterior suprasylvian areas. In addition, many of these visual areas are also interconnected with subdivisions of the dorsal lateral geniculate nucleus (LGd). Every retinotopically ordered cortical area (except ALLS and AMLS) is reciprocally interconnected with the parvocellular C layers of the LGd. The medial intralaminar nucleus of the LGd projects to areas 17, 18, 19, AMLS, and PMLS. Finally, each cortical area (except area 17) receives a projection from thalamic intralaminar nuclei. These results help to define the pathways by which visual information gains access to the vast system of extrastriate cortex in the cat.

Projections from the superior colliculus and the neocortex to the pulvinar nucleus in Galago.

We have studied the projections from the superior colliculus and the neocortex to the pulvinar nucleus in Galago senegalensis by using the retrograde transport of horseradish peroxidase (HRP). Injections of various parts of the pulvinar complex, both the inferior and superior divisions, both the tectorecipient zone and the nontectorecipient zone as defined by Glendenning et. al. ('75), produce labeled cells in the lower tier of stratum griseum superficiale. The distribution of labeled cells in the superior colliculus varies with the locus of the injection, indicating a retinotopic projection system from the entire superior colliculus to all sectors of the pulvinar complex. These experiments also provide an opportunity to study the distribution and laminar origin of neurons giving rise to cortical descending projections. The entire visual cortex projects onto the pulvinar complex. The cells or origin can be divided into two populations--one located in layer V and the other in layer VI. In seven of the nine cases reported, the layer V population is restricted entirely or mainly to the striate area. In the two exceptional cases, the layer V population is located in the adjacent extrastriate cortex, areas 18 and 19. The difference in the layer of origin of the cortical descending fibers reflects a difference in the layer of termination of the reciprocal ascending projection. These findings identify the entire visual field as primary visual cortex. The importance of this conclusion is underscored by the fact that the visual field comprises as much as one-half of the whole neocortex.

Cortical connections of the pulvinar nucleus in Galago.

In the first series of experiments, small amounts of HRP were injected into areas 17, 18, and 19 and each of the cytoarchitectonic areas of the temporal lobe. The resulting distributions of labeled cells fell into a number of distinctive classes. For example, after injecting the temporal anterior area (Ta), the labeled cells occupied a band on the ventral border of the inferior division of the pulvinar complex; after injecting the temporal ventral area (Tv), the labeled cells were concentrated in the medial extremity of the superior division. In spite of the distinctiveness of these different distributions, there was evidence that in areas that we know to contain a representation of the visual field, the injection of the same part of the field led to labeled cells in the same part of the pulvinar nucleus. Thus, when the representation of the center of the field was injected in areas 17 or 18 or Tm (the temporal middle area), labeled cells were found in the dorsal part of the rostral half of the inferior division. The distinctiveness of the different distributions did not obscure certain features common to all experiments: labeled cells were always found in both subdivisions of the pulvinar complex, and there was always continuity between the population of labeled cells in the inferior division and the population of cells in the superior division. Wherever the site of the injection in the extrastriate region, some labeled cells were found in the causal half of the inferior division. Since the caudal half of the inferior division corresponds approximately to the tecto-recipient zone as defined earlier, the entire temporal lobe, except for the auditory areas, is visual cortex. Evidence was also found for an overlap in the striate cortex between the projections of the lateral geniculate body and the pulvinar nucleus. In conclusion, the pulvinar complex projects to a vast area of cortex, including all of the occipital lobe and most of the lateral surface of the temporal lobe. This entire region, which comprises a good part of the whole of the neocortex, can be regarded as visual cortex. A second series of experiments involved cortical injections of tritiated amino acids. The results showed that the projections from the thalamus to cortex were precisely reciprocated by descending projections from cortex to thalamus. The results also served as a way of confirming the results of the first set of experiments.