Synaptic interrelationships between the optic tectum and the ipsilateral nucleus isthmi in Rana pipiens

The nucleus isthmi is reciprocally connected to the ipsilateral optic tectum. Ablation of the nucleus isthmi compromises visually guided behavior that is mediated by the tectum. In this paper, horseradish peroxidase (HRP) histochemistry and electron microscopy were used to explore the synaptic interrelationships between the optic tectum and the ipsilateral nucleus isthmi. After localized injections of HRP into the optic tectum, there are retrogradely labeled isthmotectal neurons and orthogradely labeled fibers and terminals in the ipsilateral nucleus isthmi. These terminals contain round. Clear vesicles of medium diameter (40–52 nm). These terminals make synaptic contact with dendrites of nucleus isthmi cells. Almost half of these postsynaptic dendrites are retrogradely labeled, indicating that there are monosynaptic tectoisthmotectal connections. Localized HRP injection into the nucleus isthmi labels terminals primarily in tectal layers B, E, F, and 8. The terminals contain medium-sized clear vesicles and they form synaptic contacts with tectal dendrites. There are no instances of labeled isthmotectal terminals contacting labeled dendrites. Retrogradely labeled tectoisthmal neurons are contacted by unlabeled terminals containing medium-sized and small clear vesicles. Fifty-four percent of the labeled fibers connecting the nucleus isthmi and ipsilateral tectum are myelinated fibers (average diameter approximately 0.6 μm). The remainder are unmyelinated fibers (average diameter approximately 0.4 μm). © 1994 Wiley-Liss, Inc.     

Topographic projections between the nucleus isthmi and the optic tectum in a teleost,Navodon Modestus

Following horseradish peroxidase injections into the optic tectum of a teleost,Navodon modestus, reciprocal and topographic projections between the nucleus isthmi and the ipsilateral optic tectum were determined. The isthmo-tectal fibers diverge to the optic tectum while maintaining the spatial arrangements of the isthmic cells from which the fibers originate. The tecto-isthmic projections also keep the spatial arrangements in the optic tectum. The tectal fibers converge near the nucleus isthmi and terminate in the non-cellular portion of the nucleus. The reciprocal topography is apparent in the combined results of 9 experiments with one tectal injection in each region. No labeled cells and fibers were found in the contralateral nucleus isthmi.     

Cholinergic innervation of the optic tectum in the frog Rana pipiens

An immunohistochemical method for choline acetyltransferase (ChAT) identifies presumably cholinergic axons in two retino-receptive laminae in the optic tectum of the frog Rana pipiens. Following eye enucleation there is no loss of immunoreactive axons in the optic tectum. Following unilateral ablation of the nucleus isthmi there is a near-total loss of ChAT-positive axons in the superficial cholinergic lamina contralaterally and in the deeper cholinergic lamina ipsilaterally. Thus, the cholinergic innervation of the tectum appears to derive from the nucleus isthmi. However, ChAT-positive staining of the basal optic nucleus does depend upon an intact retinal input and could derive from either retinal axons or some system trophically dependent on them.     

The nucleus isthmi as an intertectal relay for the ipsilateral oculotectal projection in the frog, Rana pipiens

Reconstruction of serial sections through the nucleus isthmi in the frog Rana pipiens shows that the cortex of the nucleus consists of a sheet of cells which is folded back on itself to form two more or less parallel faces. Cyto-architectural features allow three regions to be distinguished: a 7“rim region,” which previous work had suggested may be involved in the physiological pathway from one eye to the ipsilateral tectal lobe, and two additional regions which are here termed the anterior and posterior “nonrim cortex,” re-spectively. The cytoarchitectural features which distinguish the rim region from the other two regions are largely absent in the tadpole. Analysis of retrograde transport of horseradish peroxidase (HRP) following widespread injections into one tectal lobe indicates that the anterior nonrim cortex projects ipsilaterally while the rim cortex and posterior nonrim cortex both project contralaterally. The differing projections of the three regions are paralleled by the pattern of retrograde degeneration observed after unilateral tectal lesions. We have studied the topographic organization of the connections between the nucleus isthmi and the two tectal lobes by making small injections of HRP at tectal loci with known visual field input. Patterns of retrograde cellular labelling show that the anterior nonrim cortex projects topographically to the entirety of the tectal lobe on the same side of the brain. The rim region projects topographically to the binocularly activated part of the opposite tectal lobe; the posterior nonrim cortex projects to the monocularly activated part. There is a discontinuity in the mapping from the nucleus to the opposite tectal lobe which corresponds to the border between binocular and monocular tectum. Patterns of anterograde labelling indicate that the afferent projection from the ipsilateral tectal lobe is topographic and organized so that afferents from a given tectal locus contact cells in the rostral part of the nucleus which project back to that locus. Afferents from tectal re-gions representing binocular visual field in addition continue across the nucleus to terminate within the rim region. The organization of the afferent and efferent projections of this region is such as to link tectal loci which relate to the same direction in visual space. Our findings provide new evidence that the nucleus isthmi is involved in the pathway from one eye to the ipsilateral tectal lobe. They also show how the nucleus is organized to distribute information from a given locus in one tectal lobe to cells which project to the visually corresponding but in general anatomically different pairs of loci in the two tectal lobes. Finally, our findings confirm the existence of a projection from the nucleus isthmi to the opposite monocular tectum and suggest that this may be part of an intertectal circuit linking anatomically corresponding rather than visually corresponding pairs of tectal loci.     

Neuronal architecture of the dorsal nucleus (Cochlear nucleus) of the frog,Rana pipiens pipiens

The neuronal architecture of the dorsal nucleus of the Northern leopard frog (Rana pipiens pipiens), which is a homolog of the cochlear nucleus of mammals and birds, was investigated. Our study showed that the frog dorsal nucleus contains a number of morphologically distinct cell types that are discernible in terms of the cellular architecture as derived from Nissl-stained material and in terms of the dendritic profile as revealed by horseradish peroxidase-filled single neurons. These cell types are bushy cells, bipolar (or fusiform) cells, octopus cells, stellate cells, giant cells, radiate (or round) cells, and a variety of small cells. The different cell types occupy different regions of the nucleus. Therefore, our results suggest that the dorsal nucleus should no longer be considered to be a uniform nucleus containing a homogeneous population of neurons. Homologies of these cell types with those described in other vertebrate species, including mammals, are proposed. © 1996 Wiley-Liss, Inc.     

Musculotopic organization of the hypoglossal nucleus in the grass frog, Rana pipiens

Recent neural tracer studies in several mammalian species have demonstrated a similar musculotopic organization of the hypoglossal motoneurons which innervate individual tongue muscles. The distribution of this musculotopic organization in nonmammalian tetrapods, however, has not received detailed investigation. As part of an ongoing study on the comparative organization of the vertebrate hypoglossal nucleus, the musculotopic organization of the hypoglossal nucleus of Rana pipiens was studied by injection of lectin-conjugated horseradish peroxidase into four distinct tongue muscles and the geniohyoid muscle. Injections into the hyoglossus muscle label neurons in dorsal regions of the hypoglossal nucleus in middle and rostral nucleus levels. Injections into the genioglossus basalis muscle label neurons in ventral and lateral regions of the hypoglossal nucleus in caudal nucleus levels. Injections into the genioglossus medialis muscle label neurons in dorsal regions in caudal levels, throughout the nucleus in middle levels, and in ventral regions in more rostral levels. Injections into the geniohyoid muscle label neurons in the ventral tip of the hypoglossal nucleus and in the ventromedial corner of the medullary gray matter in middle and rostral nucleus levels. These results demonstrate that the organization of the hypoglossal nucleus in Rana pipiens is more complex than previous tracer studies indicated. Similarities in the musculotopic organization of the amphibian and mammalian hypoglossal nuclei suggest an evolutionary conservatism of the motor system controlling tongue movement.     

Putative cholinergic projections from the nucleus isthmi and the nucleus reticularis mesencephali to the optic tectum in the goldfish (Carassius auratus)

The nucleus isthmi of fish and amphibians has reciprocal connections with the optic tectum, and biochemical studies suggested that it may provide a major cholinergic input to the tectum. In goldfish, we have combined immunohistochemical staining for choline acetyltransferase with retrograde labeling of nucleus isthmi neurons after tectal injections of horseradish peroxidase. Seven fish received tectal horseradish peroxidase injections, and brain tissue from these animals was subsequently processed for the simultaneous visualization of horseradish peroxidase and choline acetyltransferase. In many nucleus isthmi neurons the dense horseradish peroxidase label obscured the choline acetyltransferase reaction product but horseradish peroxidase and choline acetyltransferase were colocalized in 54 cells from nine nuclei isthmi. The somata of nucleus reticularis mesencephali neurons stained so intensely for choline acetyltransferase that we could not determine whether they were labelled also with horseradish peroxidase. However, the large choline acetyltransferase-immunoreactive axons of nucleus reticularis mesencephali neurons stained intensely enough for us to follow them rostrally; the axons are clustered together until the level of the rostral tectum where two groupings form: one travels into the tectum and the other travels rostroventrally to cross the midline and enter the contralateral diencephalic preoptic area. We conclude therefore that cholinergic neurons project to the optic tectum from the nucleus isthmi as well as nucleus reticularis mesencephali in goldfish.     

Immunohistochemistry and spinal projections of the reticular formation in the northern leopard frog,Rana pipiens

Over 30 nuclei have been identified in the reticular formation of rats, but only a small number of distinct reticular nuclei have been recognized in frogs. We used immunohistochemistry, retrograde tracing, and cell morphology to identify nuclei within the brainstem of Rana pipiens. FluoroGold was injected into the spinal cord, and, in the same frogs, antibodies to enkephalin, substance P, somatostatin, and serotonin were localized in adjacent sections. We identified many previously unrecognized reticular nuclei. The rhombencephalic reticular formation contained reticularis (r.) dorsalis; r. ventralis, pars alpha and pars beta; r. magnocellularis; r. parvocellularis; r. gigantocellularis; r. paragigantocellularis lateralis and dorsalis; r. pontis caudalis, pars alpha and pars beta; nucleus visceralis secundarius; r. pontis oralis, pars medialis and pars lateralis; raphe obscurus; raphe pallidus; raphe magnus; and raphe pontis. The mesencephalic reticular formation contained locus coeruleus-subcoeruleus, r. cuneiformis, r. subcuneiformis, raphe dorsalis-raphe centralis superior, and raphe linearis. Thus, the reticular formation of frog, which is an anamniote, is organized complexly and is similar to the reticular formation in amniotes. Because many of these nuclei may be homologous to reticular nuclei in mammals, we used mammalian terminology for frog reticular nuclei. J. Comp. Neurol. 404:387–407, 1999. © 1999 Wiley-Liss, Inc.     

Ascending and descending projections of the lateral vestibular nucleus in the frog Rana esculenta

The lectin Phaseolus vulgaris leucoagglutinin was injected into the frog lateral vestibular nucleus (LVN) to study its antero- and retrograde projections. The following new observations were made. 1) In the diencephalon, vestibular efferents innervate the thalamus in a manner similar to that of mammalian species. The projections show a preference for the anterior, central, and ventromedial thalamic nuclei. 2) In the mesencephalon, vestibular fibers terminate in the tegmental nuclei and the nucleus of medial longitudinal fascicle. 3) In the rhombencephalon, commissural and internuclear projections interconnect the vestibular nuclei. Some of the termination areas in the reticular formation can be homologized with the mammalian inferior olive and the nucleus prepositus hypoglossi. Another part of the vestibuloreticular projection may transmit vestibular impulses toward the vegetative centers of the brainstem. A relatively weak projection is detected in the spinal nucleus of the trigeminal nerve, dorsal column nuclei, and nucleus of the solitary tract. 4) In the spinal cord, vestibular terminals are most numerous in the ipsilateral ventral horn and in the triangular area of the dorsal horn. 5) The coincidence of retrogradely labeled cells with vestibular receptive areas suggests reciprocal interconnections between these structures and the LVN. 6) In seven places, the LVN projections overlap the receptive areas of proprioceptive fibers, suggesting a convergence of sensory modalities involved in the sense of balance.     

Organization of projections from olfactory epithelium to olfactory bulb in the frog, Rana pipiens

One hypothesis for the coding of olfactory quality is that regions of the olfactory epithelium are differentially sensitive to particular odor qualities and that this regional sensitivity is conveyed to the olfactory bulb in a topographic manner by the olfactory nerve. A corollary to this hypothesis is that there is a sufficiently orderly connection between the epithelium and the olfactory bulb to convey this topographical coding. Thus we examined topography in the projection from epithelium to bulb in the frog, which has been the subject of numerous electrophysiological studies but has not yet been examined using modern neuroanatomical techniques. The tracer WGA-HRP was applied to the ventral or to the dorsal olfactory epithelium, or both. Anterograde transport of label to the olfactory bulb was seen after as few as 2 days; label was still present in the bulb as long as 21 days postinjection. In cases where WGA-HRP was applied to the entire epithelium, there was dense anterograde labelling of the ipsilateral olfactory bulb. In addition, a small medial portion of the contralateral bulb was labelled. Injections limited to either the ventral or dorsal epithelium produced patterns of anterograde labelling in the glomerular layer of the olfactory bulb, which varied with the size and location of the injection. With very large injections in either the dorsal or ventral epithelium, label appeared to be evenly distributed in the glomerular layer. With smaller injections in the ventral epithelium, there was heavier labelling in the lateral than in the medial portions of the glomerular layer, although light labelling was found in all regions of the glomerular layer. In contrast, injection sites restricted to the dorsal epithelium produced more anterograde labelling in the medial than lateral portions of the glomerular layer. These patterns extended throughout the dorsal-ventral extent of the bulb. Within the limits of the anterograde tracing technique used, we were unable to detect any systematic relationship between the pattern of labelling in the glomerular layer and the medial-lateral or rostral-caudal location of the injection site in either the ventral or dorsal epithelium. We conclude that in the frog, as in other amphibia, there is only a limited degree of topographic order between the epithelium and the olfactory bulb.     

Topographic projections of the retina and optic tectum upon the ventral lateral geniculate nucleus in the chick

The topographic projections of the retina upon the optic tectum and ventral lateral geniculate nucleus (GLv) of the chick were investigated by making small intraretinal injections of ³H-proline. The retinotectal projection pattern was similar to that described for the pigeon. The retinal projection to the GLv was also topographic and was restricted to the outermost lamina of the nucleus. The anteroposterior retinal axis was reversed in the GLv relative to its orientation in the tectum but the superoinferior axis was oriented identically in both. Furthermore, the posterior retina had an enlarged area of projection in the GLv similar to the enlarged area of retinotectal projection for the “red field” found in pigeons. The tectogeniculate projection was topographic and was confined to the outermost geniculate lamina. The secondorder retionotopic map made by the tectogeniculate projections was in register with the retinogeniculate projection. Although the retinal and tectal projection areas were coextensive in the outermost geniculate lamina, the grain density distributions peaked at different points along a radial path through the geniculate laminae. Injections of HRP into the optic tectum led to very light retrograde labeling of a small population of GLv cells topographically corresponding to the tectogeniculate projection zone of the injection site. The data suggest that the chick GLv is comparable to the GLv of other non-primate mammals.     

Some telencephalic connections in the frog, Rana pipiens.

Some afferent, efferent and intrinsic connections of the telencephalon of Rana pipiens were studied using a horseradish peroxidase method. Afferents to the telencephalon from thalamic and brain stem cell groups were demonstrated. These findings, taken together with the results of previous studies, indicate that separate thalamic cell groups project visual, auditory and somatosensory information onto the striatum. A separate thalamic cell group projects to the medial telencephalic wall and probably conveys visual and somatosensory information. These telencephalic afferent systems do not appear to be directly comparable to those of birds and reptiles. Additionally, some telencephalic afferents demonstrated in previous studies using anterograde degeneration techniques were confirmed, and some intratelencephalic connections were identified.     

Evidence for centripetally shifting terminals on the tectum of postmetamorphic Rana pipiens

In larval frogs the retina and tectum grow in topologically dissimilar patterns: new cells are added as peripheral annuli in the retina and as caudal crescents in the tectum. Retinotopy is maintained by the continual caudalward shifting of the terminals of the optic axons. After metamorphosis the pattern of growth changes. The retina continues to add new ganglion cells peripherally, but there is no neurogenesis in the tectum. To maintain retinotopy in postmetamorphic frogs, the terminals of the optic axons must continually shift toward the central tectum. We tested the proposal of centripetally shifting axons by making punctate injections of horseradish peroxidase (HRP) in the tectum of adult Rana pipiens and observing the patterns of filled cells in the contralateral retina, as was done in the goldfish (Easter and Stuermer, '84). Punctate applications of HRP in the tectum should be taken up: 1) by fascicles, and label a partial anulus of cells, 2) by terminals, and label a cluster of cells in the corresponding retinotopic site, and 3) by the extrafascicular axonal segments, and label a band of cells connecting the partial annulus to the cluster. If the terminals have shifted centripetally, the band of cells labeled through their extrafascicular segments should have a spoke-like orientation, with the center of the retina as the hub. As the tectal site moves from rostral to caudal, this band of cells should move, pendulum-like, from temporal to nasal retina. In general, the patterns of HRP-filled retinal cells we observed were consistent with our predictions. In addition, HRP taken up by the oldest (rostral) tectal axons produced more complex patterns of filled cells that indicated that these axons had shifted both caudally before metamorphosis and centripetally after.     

Neurogenesis in the optic tectum of larval Rana pipiens following unilateral enucleation

DNA synthesis and interlayer migrations of cells in the optic tectum of larval Rana pipiens were investigated, using several series of larvae which had been subjected to unilateral enucleation at stage 25, the last embryonic stage. It has been found that DNA synthesis occurs in all cellular layers of the tectum with least activity in peripheral layers. The location of the most active DNA synthesis during the larval period is the same as the location of cell division in the larval tectum, namely, in the layers bordering the ventricle of the optic lobe Unilateral enucleation of stage 25 embryos results in a decrease in DNA synthesis in all cellular layers of the optic tectum contralateral to the operation when compared to the corresponding layers in the ipsilateral tectum. The differences in rate of incorporation of ³H-thymidine between the control and affected lobes become greater during development. Fewer cells are found in each layer of the affected larval tectum than in the corresponding layer of the control tectum. The decrease is greatest in more peripheral layers, whose cells are more intimately associated with the visual circuit than are cells of the deeper layers. Peripheral to layers 1 and 2, the percentages of labeled cells found in each layer are very similar on the two sides, suggesting common factors which control migration. Differences in cell number, therefore, reflect differential cell production rather than differential cell migrations. The distribution of label resulting from ³H-thymidine incorporation at stage III indicates that the distribution of mitotic activity is not uniform in the cephalocaudal axis of the tectum. Greater cell proliferation occurs in the posterior portion of the tectum than in the anterior region throughout larval development. The peripheral control of mitotic divisions in the frog optic tectum remains unknown. The data in the present study, however, support the hypothesis that influences from afferent fibers of the optic tract modify the rates or timing of DNA synthesis in the optic tectum. The data support the notion that the deepest tectal cells respond earliest to the stimulus and these may be ependymal cells which have processes extending to the outer surface of the tectum.     

Nucleus isthmi enhances calcium influx into optic nerve fiber terminals in Rana pipiens

We examined the role of nucleus isthmi in enhancing intracellular calcium concentrations in retinotectal fibers in the frog optic tectum in vitro. The intracellular calcium levels were measured using the fluorescent calcium-sensitive dye, Calcium Green-1 3000 mw dextran conjugate (CG-1), which was injected into one optic nerve. Electrical stimulation of the labeled optic nerve alone increased tectal CG-1 fluorescence whereas electrical stimulation of nucleus isthmi alone had no effect on CG-1 fluorescence. Electrical stimulation of the nucleus isthmi ipsilateral to the labeled tectum, followed by electrical stimulation to the optic nerve can enhance calcium uptake more than a double pulse stimulation of the optic nerve alone. Maximum enhancement of the calcium signal by nucleus isthmi occurs when optic nerve stimulation follows the ipsilateral nucleus isthmi stimulation by 10 ms. These results suggest that nucleus isthmi input can facilitate retinotectal neurotransmission, and the mechanism could be used to allow the frog to attend to a single prey stimulus in an environment of several prey stimuli.     

Topographic and morphometric effects of bilateral embryonic eye removal on the optic tectum and nucleus isthmus of the leopard frog

Rana pipiens were raised through metamorphosis after extirpation of both eye primordia at Shumway embryonic stage 17 (Shumway '40). The visual connections between the isthmic nuclei and the optic tectum were examined in these animals using horseradish peroxidase (HRP) histochemistry. Isthmo-tectal projections are normally aligned with the primary retinotectal map. We asked whether these connections would develop normal topographic organization in the absence of normal retinal input. HRP was formed into a solid pellet (? 200–500 μm diameter) and inserted into one tectal lobe on the tip of a fine metal probe. The procedure produced relatively restricted retrograde label in somas and dendrites in both isthmi nuclei. In the nucleus isthmus ipsilateral to the tectal lobe receiving the HRP pellet, processes of tecto-isthmi neurons were labeled by anterograde transport. The topography of the isthmo-tectal and tecto-isthmic projections were identical in the developmentally enucleated animals and in normal frogs, even though eye removal severely reduced the volume of the optic tecta and the isthmi nuclei. Thus our analyses indicate that retinal contacts do not play an active role in the development of the positional or polarity cues that are involved in “mapping” projections between central visual nuclei. These results are discussed in the context of peripheral specification of central connections and in terms of models that have recently been proposed to explain the development of the retinotectal system.     

Retinotopic organization of central optic projections in Rana pipiens

The retinotopic organization of the anuran visual system has been investigated with the method of selective anterograde transport of horseradish peroxidase (HRP) following retinal lesions. The course of optic axons to specific structures was also confirmed by retrograde transport in the optic tract following HRP injections in the tectum and pretectum. As the optic nerve reaches the optic chiasm, the fibers from each of the four retinal quadrants appear as bands with the nasal (n) quadrant entering the chiasmal anterior pole, followed by ventral (v), temporal (t), and dorsal (d) quadrants. The preoptic nucleus is the first structure to be innervated, followed by the suprachiasmatic nucleus; both are innervated directly from fibers in the dorsal part of the optic nerve, which contains fibers from all the retinal quadrants. Each quadrant expands across the dorsoventral extent of the chiasm at the point where it enters. At this level the quadrants are arrayed along the rostrocaudal axis (as they are later in the marginal optic tract) in the sequence n-v-t-d. Optic fibers then spread across the chiasm, the nasal quadrant splits, taking up positions in the rostral and caudal margins of the optic radiation. Following the split in the nasal representation, the optic tract is transformed into topographically arranged sheets in the marginal optic tract. In the other retinorecipient nuclei, the sheet of optic axons is transformed back into the shape of the retinal hemisphere. Topographic maps of this kind display one of two possible orientations: (1) in the tectum and the nucleus lentiformis mesencephali (nLM), the temporal retina is represented in the anterior portion of the nucleus, whereas the nasal quadrant is found in the posterior portion; (2) in the thalamus, the retinotopic map is organized as a mirror-image reversal of that seen in the tectum and nLM (i.e., the nasal pole is anterior, whereas the temporal pole is in the posterior portion of the nucleus). Structures with this type of retinal map include the rostral visual nucleus, the corpus geniculatum, the nucleus of Bellonci, and the posterior thalamic nucleus. A third type of innervation occurs in the nucleus of the basal optic root (nBOR), which is the only mesencephalic visual nucleus not innervated by the marginal optic tract. The basal optic root is formed by the fibers exiting most caudally from the optic chiasm. All the retinal quadrants contribute to the basal optic root, but no evidence of retinotopy was found in nBOR.4+ target nuclei.     

The pretectal nucleus lentiformis mesencephali of Rana pipiens

The pretectal nucleus lentiformis mesencephali (nLM) of Rana pipiens was investigated with autoradiographic, horseradish peroxidase (HRP), and Golgi techniques. Retinal afferents to nLM originate primarily from the central retina. The primary projection is contralateral with a small ipsilateral component. Following optic nerve transection and HRP impregnation, contralateral retinal afferents show a restricted, dense core of HRP label in the superficial portion of the nucleus with sparser HRP label in the surround. Ipsilateral retinal afferents arborize throughout nLM, except in the dense-core region. Additional afferents to nLM originate from the ipsilateral tectum, the nucleus rotundus, the mesencephalic pretectal gray, the contralateral nLM, and the nucleus of the basal optic root. Afferents from the accessory optic system arborize only in the dense-core region, following HRP injections into the nucleus of the basal optic root, while afferents from the mesencephalic pretectal gray arborize in all parts of nLM except the dense core. Afferents from the tectum and anterior thalamus appear to arborize throughout the nucleus without discernible pattern. The lamination of afferent terminals in nLM was correlated with Nissl-stained cytoarchitectural material in which the majority of large neurons cluster around the dense core of nLM. Three types of neurons occur in nLM: large neurons (25-micron dia.), fusiform neurons (12.5-micron dia.), and stellate neurons (10-micron dia.). Additionally, two cell groups outside nLM which send dendrites into the nucleus were observed: cells of the posterior lateral nucleus and cells of the posterior thalamic pretectal gray. Both large and fusiform neurons project to the deep layers of the optic tectum as well as to the ventral rhombencephalon superficial to the abducens nucleus. While a small number of fusiform neurons project to the nucleus of the basal optic root, the stellate neurons appear to be intrinsic to nLM. The anuran nLM strongly resembles the nucleus of the optic tract in mammals in terms of the site of origin of its retinal afferents, lamination of afferent terminations, its central connections, and its demonstrated involvement in horizontal optokinetic nystagmus.