Microstructure

The neocortex essentially consists of three neuronal cell types. The most abundant are pyramidal cells. Non-pyramidal cells, also called stellate or granule cells, are divided into spiny and non-spiny neurones. All types have been subdivided on the basis of size and shape (Fig. 16.8).

Fig. 16.8 A, Characteristic neocortical neurones. Shown from left to right are the Martinotti (M), neurogliaform (N), basket (B), horizontal (H), fusiform (F), stellate (S) and pyramidal (P) types of neurone. B, The most frequent types of neocortical neurone, showing typical connections with one another and with afferent fibres (blue). Neurones limited to the cortex in their distribution are indicated in black. Efferent neurones are in magenta. The right and left afferent fibres are association or corticocortical connections; the central afferent is a specific sensory fibre. Neurones are shown in their characteristic lamina, but many have somata in more than one layer.

Pyramidal cells (Fig. 16.9) have a flask-shaped or triangular cell body ranging from 10 to 80 µm in diameter. The soma gives rise to a single thick apical dendrite and multiple basal dendrites. The apical dendrite ascends toward the cortical surface, tapering and branching, to end in a spray of terminal twigs in the most superficial lamina, the molecular layer. From the basal surface of the cell body, dendrites spread more horizontally, for distances up to 1 mm for the largest pyramidal cells. Like the apical dendrite, the basal dendrites branch profusely along their length. All pyramidal cell dendrites are studded with myriad dendritic spines. These become more numerous as the distance from the parent cell soma increases. A single slender axon arises from the axon hillock, which is usually situated centrally on the basal surface of the pyramidal neurone. Ultimately, in the vast majority of (if not all) cases, the axon leaves the cortical grey matter to enter the white matter. Pyramidal cells are thus, perhaps universally, projection neurones. They appear to use excitatory amino acids, either glutamate or aspartate, exclusively as their neurotransmitters.

Fig. 16.9 Neurones in the cerebral cortex. A, A single pyramidal cell stands out among many unstained elements. B, Isolated Golgi-stained neurones are prominent among the Nissl-stained cortical elements.

(Preparations provided by A. R. Lieberman, Department of Anatomy, University College, London.)

Spiny stellate cells are the second most numerous cell type in the neocortex and, for the most part, occupy lamina IV. They have relatively small multipolar cell bodies, commonly 6 to 10 µm in diameter. Several primary dendrites, profusely covered in spines, radiate for varying distances from the cell body. Their axons ramify within the grey matter, predominantly in the vertical plane. Spiny stellate cells are likely to use glutamate as their neurotransmitter.

The smallest group comprises the heterogeneous non-spiny or sparsely spinous stellate cells. All are interneurones, and their axons are confined to grey matter. In morphological terms, this is not a single class of cell but a multitude of different forms, including basket, chandelier, double bouquet, neurogliaform, bipolar-fusiform and horizontal cells. Various types may have horizontally, vertically or radially ramifying axons.

Neurones with mainly horizontally dispersed axons include basket and horizontal cells. Basket cells have a short, vertical axon that rapidly divides into horizontal collaterals; these end in large terminal sprays synapsing with the somata and proximal dendrites of pyramidal cells. The cell bodies of horizontal cells lie mainly at the superficial border of lamina II and occasionally deep in lamina I (the molecular or plexiform layer). They are small and fusiform, and their dendrites spread short distances in two opposite directions in lamina I. Their axons often stem from a dendrite, then divide into two branches that travel away from each other for great distances in the same layer.

Neurones with an axonal arborization that is predominantly perpendicular to the pial surface include chandelier, double bouquet and bipolar-fusiform cells. Chandelier cells have a variable morphology, although most are ovoid or fusiform, and their dendrites arise from the upper and lower poles of the cell body. The axonal arborization, which emerges from the cell body or a proximal dendrite, is characteristic and identifies these neurones. A few cells in the more superficial laminae (II and IIIa) have descending axons; deeper cells (laminae IIIc and IV) have ascending axons, and intermediate neurones (IIIb) often have both. The axons ramify close to the parent cell body and terminate in numerous vertically oriented strings that run alongside the axon hillocks of the pyramidal cells with which they synapse. Double bouquet (or bitufted) cells are found in laminae II and III, and their axons traverse laminae II and V. Generally, these neurones have two or three main dendrites that give rise to superficial and deep dendritic tufts. A single axon usually arises from the oval or spindle-shaped cell soma and rapidly divides into ascending and descending branches. These branches collateralize extensively, but the axonal arbor is confined to a perpendicularly extended yet horizontally confined cylinder, 50 to 80 µm across. Bipolar cells are ovoid, with a single ascending dendrite and a single descending dendrite that arise from the upper and lower poles, respectively. These primary dendrites branch sparsely. Their branches run vertically to produce a narrow dendritic tree, rarely more than 100 µm across, which may extend through most of the cortical thickness. Commonly, the axon originates from one of the primary dendrites and rapidly branches to form a vertically elongated, horizontally confined axonal arbor that closely parallels the dendritic tree in extent.

The principal recognizable neuronal type is the neurogliaform or spiderweb cell. These small spherical cells, 10 to 12 µm in diameter, are found mainly in laminae II to IV, depending on the cortical area. Seven to 10 thin dendrites typically radiate from the cell soma, some branching once or twice to form a spherical dendritic field measuring 100 to 150 µm in diameter. The slender axon arises from the cell body or a proximal dendrite. Almost immediately, it branches profusely within the vicinity of the dendritic field (and usually somewhat beyond it), to form a spherical axonal arbor up to 350 µm in diameter.

The majority of non-spiny or sparsely spinous non-pyramidal cells probably use γ-aminobutyric acid (GABA) as their principal neurotransmitter. This is almost certainly the case for basket, chandelier, double bouquet, neurogliaform and bipolar cells. Some are also characterized by the coexistence of one or more neuropeptides, including neuropeptide Y, vasoactive intestinal polypeptide (VIP), cholecystokinin, somatostatin and substance P. Acetylcholine is present in a subpopulation of bipolar cells, which may also be GABAergic and contain VIP.

Laminar Organization

The most apparent microscopic feature of the neocortex stained for cell bodies or for fibres is its horizontal lamination. Its value for understanding cortical functional organization is debatable, but the use of cytoarchitectonic descriptions to identify regions of cortex is common.

Typical neocortex is described as having six layers or laminae lying parallel to the surface (Figs 16.10–16.12).

Fig. 16.10 Layers of the cerebral cortex. The three vertical columns represent the disposition of cellular elements as revealed by the staining techniques of Golgi (impregnating whole neurones), Nissl (staining cell bodies) and Weigert (staining nerve fibres).

Fig. 16.11 A and B, Distribution and characteristics of the five major types of cerebral cortex.

Fig. 16.12 Lateral (A) and medial (B) surfaces of the left cerebral hemisphere depicting Brodmann’s areas.

Neocortical Structure

Five regional variations are described in neocortical structure (see Fig. 16.11). Although all are said to develop from the same six-layered pattern, two types—granular and agranular—are regarded as virtually lacking certain laminae and are referred to as heterotypical. Homotypical variants, in which all six laminae are found, are called frontal, parietal and polar—names that link them with specific cortical regions in a somewhat misleading manner (e.g. the frontal type also occurs in parietal and temporal lobes).

The agranular type has diminished or absent granular laminae (II and IV) but always contains scattered stellate somata. Large pyramidal neurones are found in the greatest densities in agranular cortex, which is typified by the numerous efferent projections of pyramidal cell axons. Although it is often equated with motor cortical areas such as the precentral gyrus (area 4), agranular cortex also occurs elsewhere, such as in areas 6, 8 and 44 and parts of the limbic system.

In the granular type of cortex, the granular layers are maximally developed and contain densely packed stellate cells, among which small pyramidal neurones are dispersed. Laminae III and IV are poorly developed or unidentifiable. This type of cortex is particularly associated with afferent projections. However, it does have efferent fibres, derived from the scattered pyramidal cells, although they are less numerous than elsewhere. Granular cortex occurs in the postcentral gyrus (somatosensory area), striate area (visual area) and superior temporal gyrus (acoustic area) and in small parts of the parahippocampal gyrus. Despite its very high density of stellate cells, especially in the striate area, it is the second thinnest of the five main types. In the striate cortex, the external band of Baillarger (lamina IV) is well defined as the stria (white line) of Gennari.

The other three types of cortex are intermediate forms. In the frontal type, large numbers of small and medium-sized pyramidal neurones appear in laminae III and V, and the granular layers (II and IV) are less prominent. The relative prominence of these major forms of neurone vary reciprocally wherever this form of cortex exists.

The parietal type of cortex contains pyramidal cells, which are mostly smaller than in the frontal type. The granular laminae are, on the contrary, wider and contain more stellate cells. This kind of cortex occupies large areas in the parietal and temporal lobes. The polar type is classically identified with small areas near the frontal and occipital poles. It is the thinnest form of cortex. All six laminae are represented, but the pyramidal layer (III) is reduced in thickness and not as extensively invaded by stellate cells as in the granular type of cortex. In both polar and granular types, the multiform layer (VI) is more highly organized than in other types.

For almost 100 years it has been customary to refer to discrete cortical territories not only by their anatomical location in relation to gyri and sulci but also in relation to their cytoarchitectonic characteristics as originally descried by Brodmann (see Fig. 16.12). Some of the areas so defined (e.g. the primary sensory and motor cortices) have clear relevance in terms of anatomical connections and functional significance; others less so.

Overview of Cortical Connectivity

All neocortical areas have axonal connections with other cortical areas on the same side (ipsilateral corticocortical or association connections) and on the opposite side (contralateral corticocortical or commissural connections) and with subcortical structures.

The primary somatosensory, visual and auditory areas give rise to ipsilateral corticocortical connections to the association areas of the parietal, occipital and temporal lobes, respectively, which then progressively project toward the medial temporal limbic areas, notably the parahippocampal gyrus, entorhinal cortex and hippocampus. Thus, the first somatosensory area (SI) projects to the superior parietal cortex (Brodmann’s area 5), which in turn projects to the inferior parietal cortex (area 7). From there, connections pass to cortex in the walls of the superior temporal sulcus, to the posterior parahippocampal gyrus and into limbic cortex. Similarly, the primary visual cortex (area 17) projects to the parastriate cortex (area 18), which in turn projects to the peristriate region (area 19). Information then flows to inferotemporal cortex (area 20), to cortex in the walls of the superior temporal sulcus, to medial temporal cortex in the posterior parahippocampal gyrus and to limbic areas. The auditory system shows a similar progression from primary auditory cortex to temporal association cortex and finally to the medial temporal lobe.

In addition to this stepwise outward progression from sensory areas through posterior association cortex, connections occur at each stage with parts of the frontal cortex. Thus, taking the somatosensory system as an example, primary somatosensory cortex (SI) in the postcentral gyrus is reciprocally connected with primary motor cortex (area 4) in the precentral gyrus. As the next step in the outward progression, the superior parietal lobule (area 5) is interconnected with the premotor cortex (area 6), which in turn is connected with area 7 in the inferior parietal lobule. This has reciprocal connections with prefrontal association cortex on the lateral surface of the hemisphere (areas 9 and 46) and with temporal association areas, which connect with more anterior prefrontal association areas and, ultimately, with orbitofrontal cortex. Similar stepwise links exist between areas on the visual and auditory association pathways in the occipitotemporal lobe and areas of the frontal association cortex. The connections between sensory and association areas are reciprocal.

All neocortical areas are connected with subcortical regions, although the density of these connections varies among areas. First among these are connections with the thalamus (Ch. 15). All areas of the neocortex receive afferents from more than one thalamic nucleus, and all such connections are reciprocal. The vast majority of (if not all) cortical areas project to the striatum, tectum, pons and brain stem reticular formation. Additionally, all cortical areas are reciprocally connected with the claustrum; the frontal cortex connects with the anterior part, and the occipital lobe with the posterior part.

All cortical areas receive a topographically organized cholinergic projection from the basal forebrain, which is profoundly affected by the neurodegenerative processes of Alzheimer’s disease. Similarly, noradrenergic fibres pass to all cortical areas from the locus coeruleus, as do serotoninergic fibres from the midbrain raphe nuclei, histaminergic fibres from the posterior hypothalamus and dopaminergic fibres from the ventral midbrain.

Different cortical areas have widely different afferent and efferent connections. Some have connections that are unique, such as the corticospinal motor projection (corticospinal tract) from pyramidal cells in a restricted area around the central sulcus.

Widely separated but functionally interconnected areas of cortex share common patterns of connections with subcortical nuclei and within the neocortex. For example, contiguous zones of the striatum, thalamus, claustrum, cholinergic basal forebrain, superior colliculus and pontine nuclei connect with anatomically widely separated areas in the prefrontal and parietal cortices, which are themselves interconnected. In contrast, other cortical regions that are functionally distinct (e.g. areas in the temporal and parietal cortices) do not share such contiguity in their subcortical connections. (See Case 12.)

Frontal Lobe

The frontal lobe is the rostral region of the hemisphere, anterior to the central sulcus and above the lateral fissure. On the superolateral surface, extending onto the medial surface, is the precentral gyrus, running parallel to the central sulcus and limited anteriorly by the precentral sulcus. The area of the frontal lobe anterior to the precentral sulcus is divided into the superior, middle and inferior frontal gyri (see Figs 16.1, 16.2). In front of these gyri lies the frontal pole. The ventral surface of the frontal lobe overlies the bony orbit and is the orbitofrontal cortex. The medial surface extends from the frontal pole anteriorly to the paracentral lobule behind. It consists of the medial frontal cortex and the anterior cingulate cortex.

Primary Motor Cortex

The primary motor cortex (MI) corresponds to the precentral gyrus (area 4). It is the area of cortex with the lowest threshold for eliciting contralateral muscle contraction by electrical stimulation. The primary motor cortex contains a detailed, topographically organized map (motor homunculus) of the opposite body half, with the head represented most laterally and the legs and feet represented on the medial surface of the hemisphere in the paracentral lobule (Fig. 16.13). A striking feature is the disproportionate representation of body parts in relation to their physical size. Thus, large areas represent the muscles of the hand and face, which are capable of finely controlled or fractionated movements.

Fig. 16.13 A, The motor homunculus showing proportional somatotopic representation in the main motor area. B, The sensory homunculus showing proportional somatotopic representation in the somaesthetic cortex.

(After Penfield, W., Rasmussen, T., 1950. The Cerebral Cortex of Man. Macmillan, New York.)

The cortex of area 4 is agranular, and layers II and IV are difficult to identify. The most characteristic feature is the presence in lamina V of some extremely large pyramidal cell bodies, Betz cells, which may approach 80 µm in diameter. The axons of these neurones project into the corticospinal and corticobulbar tracts.

The major thalamic connections of area 4 are with the ventral posterolateral nucleus, which in turn receives afferents from the deep cerebellar nuclei. The ventral posterolateral nucleus also contains a topographical representation of the contralateral body, which is preserved in its point-to-point projection to area 4, where it terminates largely in lamina IV. Other thalamic connections of area 4 are with the centromedian and parafascicular nuclei. These seem to provide the only route through which output from the basal ganglia, via the thalamus, reaches the motor cortex; this appears to be the case, because the projection of the internal segment of the globus pallidus to the ventrolateral nucleus of the thalamus is confined to the anterior division, and there is no overlap with cerebellothalamic territory. The anterior part of the ventrolateral nucleus projects to the premotor and supplementary motor areas of cortex, with no projection to area 4.

The ipsilateral somatosensory cortex (SI) projects in a topographically organized way to area 4, and the connection is reciprocal. The projection to the motor cortex arises in areas 1 and 2, with little or no contribution from area 3b. Fibres from SI terminate in layers II and III of area 4, where they contact mainly pyramidal neurones. Evidence suggests that neurones activated monosynaptically by fibres from SI, as well as those activated polysynaptically, make contact with layer V pyramidal cells, which give rise to corticospinal fibres, including Betz cells. Movement-related neurones in the motor cortex that can be activated from SI tend to have a late onset of activity, mainly during the execution of movement. It has been suggested that this pathway plays a role primarily in the making of motor adjustments during a movement. Additional ipsilateral corticocortical fibres to area 4 from behind the central sulcus come from the second somatosensory area (SII).

Neurones in area 4 are responsive to peripheral stimulation and have receptive fields similar to those in the primary sensory cortex. Cells located posteriorly in the motor cortex have cutaneous receptive fields, whereas more anteriorly situated neurones respond to stimulation of deep tissues.

The motor cortex receives major frontal lobe association fibres from the premotor cortex and the supplementary motor area, as well as fibres from the insula. It is probable that these pathways modulate motor cortical activity in relation to the preparation, guidance and temporal organization of movements. Area 4 sends fibres to, and receives fibres from, its contralateral counterpart and also projects to the contralateral supplementary motor cortex.

Apart from its contribution to the corticospinal tract, the motor cortex has diverse subcortical projections. The connections to the striatum and pontine nuclei are heavy. It also projects to the subthalamic nucleus. The motor cortex sends projections to all nuclei in the brain stem, which are themselves the origin of descending pathways to the spinal cord—namely, the reticular formation, red nucleus, superior colliculus, vestibular nuclei and inferior olivary nucleus.

Corticospinal Tract

The corticospinal or pyramidal tract provides direct control by the cerebral cortex over motor centres of the spinal cord. A homologous pathway to the brain stem, the corticobulbar or corticonuclear projection, fulfills a similar function in relation to motor nuclei of the brain stem. The corticospinal tract does not originate solely from the motor cortex but is conveniently considered in conjuction with it.

The percentage of corticospinal fibres that arise from the primary motor cortex may actually be quite small, probably in the range of 20% to 30%. They arise from pyramidal cells in layer V and give rise to the largest-diameter corticospinal axons. There is also a widespread origin from other parts of the frontal lobe, including the premotor cortex and the supplementary motor area. Many axons from the frontal cortex, notably the motor cortex, terminate in the ventral horn of the spinal cord. In cord segments mediating dexterous hand and finger movements, they terminate in the lateral part of the ventral horn, in close relationship to motor neuronal groups. A small percentage establish direct monosynaptic connections with α motor neurones.

Between 40% and 60% of pyramidal tract axons arise from parietal areas, including area 3a, area 5 of the superior parietal lobe, and SII in the parietal operculum. The majority of parietal fibres to the spinal cord terminate in the deeper layers of the dorsal horn.

Corticomotor neuronal cells are active in relation to agonist muscle force of contraction; their relation to amplitude of movement is less clear. Their activity precedes the onset of electromyographic activity by 50 to 100 milliseconds, suggesting a role for cortical activation in generating rather than monitoring movement.

Premotor Cortex

Immediately in front of the primary motor cortex lies Brodmann’s area 6 (Fig. 16.14). Area 6 extends onto the medial surface, where it becomes contiguous with area 24 in the cingulate gyrus, anterior and inferior to the paracentral lobule. A number of functional motor areas are contained in this cortical region. Lateral area 6, the area over most of the lateral surface of the hemisphere, corresponds to the premotor cortex. The premotor cortex is divided into dorsal and ventral areas on functional grounds and on the basis of ipsilateral corticocortical association connections.

Fig. 16.14 Lateral (A) and medial (B) surfaces of the left cerebral hemisphere showing the approximate correspondence of Brodmann’s areas to the primary motor cortex (area 4), premotor areas (areas 6 and 8) and motor speech areas (areas 44 and 45).

The major thalamic connections of the premotor cortex are with the anterior division of the ventrolateral nucleus and with the centromedian, parafascicular and centrolateral components of the intralaminar nuclei. Subcortical projections to the striatum and pontine nuclei are prominent, and this area also projects to the superior colliculus and the reticular formation. Both dorsal and ventral areas contribute to the corticospinal tract. Commissural connections are with the contralateral premotor, motor and superior parietal (area 5) cortices. Ipsilateral corticocortical (association) connections with area 5 in the superior parietal cortex and inferior parietal area 7b are common to both dorsal and ventral subdivisions of the premotor cortex, and both send a major projection to the primary motor cortex. The dorsal premotor area also receives fibres from the posterior superior temporal cortex and projects to the supplementary motor cortex. The frontal eye field (area 8) projects to the dorsal subdivision. Perhaps the greatest functionally significant difference in connectivity between the two premotor area subdivisions is that the dorsal premotor area receives from the dorsolateral prefrontal cortex, whereas the ventral subdivision receives from the ventrolateral prefrontal cortex. All these association connections are either likely or known to be reciprocal.

Neuronal activity in the premotor cortex in relation to both preparation for movement and movement itself has been extensively studied experimentally. Direction selectivity for movement is a common feature of many neurones. In behavioural tasks, neurones in the dorsal premotor cortex show anticipatory activity and task-related discharge as well as direction selectivity, but little or no stimulus-related changes. The dorsal premotor cortex is probably important in establishing a motor set or intention, contributing to motor preparation in relation to internally guided movement. In contrast, ventral premotor cortex is related more to the execution of externally (especially visually) guided movements in relation to a specific external stimulus.

Frontal Eye Field

The frontal eye field lies predominantly within Brodmann’s area 8, anterior to the superior premotor cortex (Fig. 16.15). It receives its major thalamic projection from the parvocellular mediodorsal nucleus, with additional afferents from the medial pulvinar, the ventral anterior nucleus and the suprageniculate–limitans complex. It connects with the paracentral nucleus of the intralaminar group. The thalamocortical pathways to the frontal eye field form part of a pathway from the superior colliculus, the substantia nigra and the dentate nucleus of the cerebellum. The frontal eye field has extensive ipsilateral corticocortical connections, receiving fibres from several visual areas in the occipital, parietal and temporal lobes, including the medial temporal area (V5) and area 7a. There is also a projection from the superior temporal gyrus, which is auditory rather than visual in function. From within the frontal lobe, the frontal eye field receives fibres from the ventrolateral and dorsolateral prefrontal cortices. It projects to the dorsal and ventral premotor cortices and to the medial motor area, probably to the supplementary eye field adjacent to the supplementary motor area proper. It projects prominently to the superior colliculus, the pontine gaze centre within the pontine reticular formation and other oculomotor-related nuclei in the brain stem. As its name implies, it is important in the control of eye movements. Destructive lesions of the frontal eye field cause ipsilateral conjugate deviation of the eyes, whereas stimulation, such as with an epileptic discharge, induces contralateral deviation.

Fig. 16.15 Lateral surface of the left cerebral hemisphere showing the frontal eye field, corresponding to parts of Brodmann’s areas 6, 8 and 9. The perimeter of this area is delineated by an interrupted line to indicate the uncertainty of its precise extent.

CASE 1 Gaze Deviation

A 75-year-old, right-handed man with a long history of hypertension and hypercholesterolemia is admitted to the hospital after awakening with left-sided weakness. He is mildly confused but awake and able to answer questions. The visual fields are full to confrontation. Head and eyes are deviated to the right, and he has a left hemiplegia and left hemihypaesthesia. Magnetic resonance imaging (MRI) shows a large ischaemic infarct in the territory of the right middle cerebral artery.

Subsequent examinations continue to document head and eye deviations to the right. The vestibulo-ocular (‘doll’s eye’) reflex induces only very transient eye movements to the left of midline. Four days after admission, the head deviation has improved significantly, as has spontaneous eye movement to the left of the midline. Thereafter, the patient is increasingly able to direct his gaze to the left.

Discussion: Acute destructive (ischaemic) cerebral hemispheric lesions involving the frontal (‘adversive’) eye field are well known to produce ipsilateral eye deviation, with a markedly impaired ability to direct voluntary gaze past the midline to the opposite side on command. Over a period of several days, the gaze deviation generally improves and ultimately resolves. This is in contrast to an irritative lesion (e.g. a focal seizure), which induces deviation of the head and eyes to the opposite side.

Parietal Lobe

The parietal lobe lies posterior to the central sulcus. On the medial aspect of the hemisphere, its boundary with the occipital lobe is clearly demarcated by the deep parieto-occipital sulcus. On the lateral aspect of the hemisphere, its boundaries with the occipital and temporal lobes are less distinct and somewhat arbitrary. The inferior boundary is the posterior ramus of the lateral fissure and its imaginary posterior prolongation.

The lateral aspect of the parietal lobe is divided into three areas by postcentral and intraparietal sulci (see Fig. 16.1). The postcentral sulcus, often divided into upper and lower parts, is posterior and parallel to the central sulcus. Inferiorly, it ends above the posterior ramus of the lateral fissure. The postcentral gyrus or primary somatosensory cortex lies between the central and postcentral sulci. Posterior to the postcentral sulcus there is a large area, subdivided by the intraparietal sulcus. It usually starts in the postcentral sulcus near its midpoint and extends posteroinferiorly across the parietal lobe, dividing it into superior and inferior parietal lobules. Posteriorly, its occipital ramus extends into the occipital lobe, joining the transverse occipital sulcus at right angles.

The superior parietal lobule, between the superomedial margin of the hemisphere and the intraparietal sulcus, is continuous anteriorly with the postcentral gyrus around the upper end of the postcentral sulcus; posteriorly, it often joins the arcus parieto-occipitalis, surrounding the lateral part of the parieto-occipital sulcus.

The inferior parietal lobule, below the intraparietal sulcus and behind the lower part of the postcentral sulcus, is divided into three parts. The anterior part is the supramarginal gyrus, which arches over the upturned end of the lateral fissure. It is continuous anteriorly with the lower part of the postcentral gyrus and posteroinferiorly with the superior temporal gyrus. The middle part of the inferior parietal lobule, called the angular gyrus, arches over the end of the superior temporal sulcus and is continuous posteroinferiorly with the middle temporal gyrus. The posterior part of the inferior parietal lobule arches over the upturned end of the inferior temporal sulcus onto the occipital lobe, forming an arcus temporo-occipitalis.