Identification of synapses formed in the cerebellar cortex by purkinje axon collaterals: an electron microscope study

Summary Purkinje axon collaterals and their synaptic terminals can be identified on the basis of three criteria: (1) They are the only myelinated axons of local elements, hence any myelinated axon persisting in chronically isolated folium is a Purkinje axon or its collateral; (2) They are the only known transfolial axons, so that axons and synapses found in the state of secondary degeneration after lesions placed into neighbouring folia of the cerebellar cortex are Purkinje axon collaterals and synapses; (3) The peculiar axonal tubular systems described by Andres (1965) are specific for Purkinje axons and their synaptic endings, which offers an additional clue for their identification. Using these three criteria numerous synapses of Purkinje axon collateral endings have been identified on the large Golgi neurons, both cell bodies and principal dendrites, and on the bodies of basket neurons. No evidence of the termination of Purkinje axon collaterals on other Purkinje cells could be detected.     

Correlations among climbing fiber responses of nearby cerebellar Purkinje cells in the immature rat

Summary Correlations between pairs of spontaneous climbing fiber responses (CFRs) recorded from couples of nearby Purkinje cells (PCs) were studied in immature rats by using cross-correlograms between CFR pairs, and compared to those in adult animals. Correlations were found as early as day 3. Some days later, on PN days 7–9, these correlations were higher than in the adult. In most cases, this was apparently not due to the multiple innervation of PC by climbing fibers (CFs) which normally occurs during this immature stage since: 1) temporal relationships between the paired CFRs varied by more than 30 ms and 2) thresholds for pairs of graded CFRs and additional components of the responses evoked in the 2 PCs by juxtafastigial or olivary stimulation were different. Synchronizing mechanisms were therefore likely to be already located at the olivary level. However, in 3 couples of multiply innervated PCs whose spontaneous CF activities were highly correlated, stimulation experiments revealed a common innervation of the 2 cells by branches of the same CF. In multiply innervated cells, spontaneous responses mediated through distinct CFs were also synchronized, suggesting that these fibers originate from neighboring neurons of the inferior olive. Finally, in 7 to 9-day-old rats, correlations among CFR pairs were much more restricted in the longitudinal axis of the folium than in the transverse one. On the whole, the present study indicates that correlations among CFRs of nearby PCs exist as soon as CF-PC synapses are established and the latter are already organized in sagittal strips at early stages of development.This work was supported by INSERM (C.R.L. Nr. 78.1.003.6)     

The distribution to the cerebellar anterior lobe of the climbing and mossy fiber inputs from the plantar and palmar cutaneous afferents

Summary Systematic examination has been made of the potentials evoked in the ipsilateral anterior lobe by single Group II volleys in different branches of cutaneous nerves to the fore-paw and hind-paw of the cat. Field potentials evoked by the mossy and climbing fiber inputs have been recorded along microelectrode tracks arranged so that there has been a comprehensive study through the whole branching foliated structure. In a previous investigation it was shown that large cutaneous nerves of the forelimb and hindlimb have wide fields of action for both the mossy fiber and climbing fiber inputs. In this present investigation it was found that small cutaneous nerves have more localized distributions within these wide fields. This discriminative distribution is exhibited for Group II volleys in the subdivisions of the nerves providing innervation to the palmar and plantar foot pads. It thus appears from this somatotopic investigation that there are pathways to the cerebellum sufficiently specific to give information about the part of the foot that is being stimulated in natural movements.     

The climbing fibers of the cerebellar cortex,their origin and pathways in cat

Summary The sources and pathways of the climbing fibers to the cerebellar posterior vermis were studied with combined electrophysiological and anatomical methods in cats.Recording from identified cerebellar Purkinje cells, monosynaptic climbing fiber (CF) responses have been obtained both for stimulation of the inferior olive (IO) and various parts of the brain stem (BS). CF responses were found to be of three types, IO only, BS only or both IO and BS. However the responses to BS stimulation were very few in number in comparison with IO or IO and BS types of responses. The latencies of the responses were shorter for the BS cases consistent with their distance from the cerebellum.A comparison of latencies and the relative responsiveness of the different area of the brain stem which were studied, indicate that part of the CF ascend through the pontine region and enter the cerebellum by way of the medium and superior peduncles. This finding is confirmed by the results of anatomical studies in which degenerating fibers were found in the molecular layer (using the Nauta technique) after lesion of the brachium pontis but not after lesions of the medial portion of the pons. Similarly, injection of radioactive leucine into the pontine nuclei failed to show any labeled fibers in the molecular layer.Horseradish peroxidase (HRP) was injected into localized regions of the posterior vermis after total bilateral destruction of the inferior peduncles. Large numbers of positive, marked cells were still found in the inferior olive.It is concluded that nearly all, if not all, the climbing fibers originate in the inferior olive and that they ascend to the cerebellum by way of all the peduncles.     

Early terminal degeneration of cerebellar climbing fibers after destruction of the inferior oliver in the rat. Synaptic relationships in the molecular layer

Summary The cerebellar molecular layer in adult rats has been studied with the electron microscope at several early and consecutive survival times following 3-acetylpyridine intoxication. Climbing fiber (CF) terminals underwent a fast process of electron-dense degeneration which became apparent from 16 hours onwards. A small proportion of degenerating terminals were depleted of vesicles and filled with a dark flocculent and granular homogeneous matrix. Microtubular changes in degenerating CF tendrils were observed. CF terminals were found in relation with every Purkinje cell in normal animals and completely disappeared within 72 hours after the treatment. CF synapses were found on Purkinje dendritic and somatic thorns, sometimes also on the dendritic shafts or even on the Purkinje soma. Convincing evidence of synaptic contacts of CF varicosities on either basket or stellate cells could not be obtained. CF synapses with Golgi II cell dendrites in the molecular layer were described. Decrease in the number of post-synaptic dendritic thorns normally assigned fo CF synapses was observed consequential to CF anterograde degeneration. The observations are consistent with previous conclusions drawn from light microscopic studies that the clearing up of CF debris in the molecular layer is completed within the short time of three days, and that the inferior olive seems to be the only source of CFs.     

Maturation of evoked climbing fiber input to rat cerebellar Purkinje cells (I.)

Summary An analysis of evoked responses of Purkinje cells in developing rat cerebellum to climbing fiber input was conducted to determine which identifying properties of this afferent system are established early in development and which specific features mature with age. Rat pups at various ages were anesthetized with 0.5% halothane and unit recordings made with glass micropipettes. By the third postnatal day, electrical stimulation of the sensorimotor cortex and limbs at low stimulation rates (<1/sec) could elicit distinct burst responses appearing at long latencies (180 msec), indicating that pathways of both ascending and descending climbing fiber systems are intact early in cerebellar cortical development. A distinctive feature maturing over the first 1–1.5 weeks was the characteristic of the all-or-none burst response since before about day 11 the interspike interval, amount of inactivation, and number of spikes in evoked burst responses all varied from stimulation to stimulation. Mean latencies decreased from 180 msec at day 3 to 50 msec by day 10, but did not achieve the adult value of 20 msec until the fourth week. Typically, climbing fiber responses could only follow at stimulation rates of less than 0.2/sec at day 3, but by day 12 could follow up to 10/sec, which is the same as in the adult. The data indicate that the climbing fiber system establishes connectivity from diverse sources and exhibits identifying characteristics similar to the adult early in cerebellar development. Most aspects of the maturation of transmission can be explained if there were a decrease in the time scale of function of the synapses involved, mainly those in the inferior olive.Supported by N.I.H. Grant 5-R01-6M00133 and N.S.F. Grant GB 43301     

Participation of Golgi neuron processes in the cerebellar glomeruli: An electron microscope study

Summary Synaptic relations, within the cerebellar isles, of Golgi II neuron axons and dendrites have been studied in the cat. Golgi axon endings can be identified with some probability in the outer (cortical) zone of the cerebellar glomeruli in normal material. They can well be recognized in the chronically isolated folium in which mossy fibers have completely degenerated. The Golgi axons are very thin preterminal fibers with small enlargements containing synaptic vesicles and contacting the preterminal intraglomerular parts of the granule cell dendrites as well as their terminal spheroid protrusions. The spheroid protrusions of the granule dendrites are the main postsynaptic loci of the granule neuron having their main synapse with the mossy fiber — generally of central position in the glomerulus — and additional synapses, more often on their outer surface, with the Golgi axons. No significant difference is seen between the two contacts, from which one is known to be excitatory (mossy) and the other inhibitory (Golgi ax.). The Golgi cell has also descending dendrites, known from light microscopy to be engaged in the cerebellar isles. By tracing these dendrites from the cell bodies and using their characteristic short finger-like processes as a criterion for their identification, the synapses between mossy endings and Golgi dendrites could be identified under the EM. They are broad contacts between a dendrite passing along one side of the mossy ending, with several synaptic attachment plaques and with small dendritic processes protruding into invaginations of the mossy ending. — The cerebellar glomerulus is thus a complex synaptic apparatus with two different axonal elements (mossy and Golgi endings) and often two dendritic elements (granule and Golgi dendrites) involved. — The possible functional significance of the Golgi cell is discussed in the light of these findings and the new discoveries by Eccles et al. (1964b, 1966) on its inhibitory nature.     

Synaptic organization of the Purkinje cells in the cerebellum of the cat

Summary An attempt at distinction between excitatory and inhibitory synapses is made in the cat cerebellum. The former are assumed to contain spheroid vesicles (S-type) of average diameter of 500 Å, while the latter flattened vesicles (F-type) of smaller size than the former. The elongation index (the ratio of the length of major versus minor axis of the vesicles) of S-type synaptic vesicles was about 1.2, while that of the F-type was more than 1.7. Parallel fibers of granule cells make S-type synaptic contacts (en-passant type or crossing-over synapse) mostly on the spines of the smaller branchlets of Purkinje cells. Climbing fibers make also S-type synapses on the smaller spines with short necks of the larger dendrites of Purkinje cells, but not frequently on the direct surface of them. It must be emphasized that almost no F-type synapse has been recognized which makes synaptic contacts directly on the spine of any type. It makes synaptic contacts usually on the direct surface of dendrites of Purkinje cells. Basket cell axons embrace directly the somas of the Purkinje cells. Their synaptic contacts were always of F-type and of en-passant character.The hypothesis is proposed that excitatory (E-type) synapses can be identified with synapses of S-type, whereas inhibitory (I-type) synapses would correspond to the F-type terminals.     

Organization of climbing fiber input from mechanoreceptors to lobule V vermal cortex of the cat

Summary The somatotopic organization of the climbing fiber (CF) projections to the vermal cortex of lobule V of the cat was revealed by low threshold natural stimulation of mechanoreceptors. Extracellular single-unit recordings were made from 554 Purkinje cells in cats anesthetized with sodium pentobarbital. Forty-nine percent of the CF responses were elicited by cutaneous stimulation of the forelimb (62%), hindlimb (25%), or upper back and neck (13%). The topographical arrangement consisted of a 1 mm wide medial zone and a 1–1.5 mm wide lateral zone. In the medial zone, the CF responses were mainly nonresponsive to any cutaneous stimulation except in the caudomedial portion of the lobule where the upper back, neck or ears were represented in a narrow parasagittally oriented strip. The lateral zone contained a mixture of CF responses representing projections from different portions of the ipsilateral forelimb and hindlimb. Although CF responses connected with the forepaw or hindpaw predominated throughout all parts of the lateral zone, the more medial portions of this zone contained larger receptive fields involving the more proximal areas of the limb whereas the lateral part of the zone had smaller receptive fields representing the distal regions, particularly the ventral forepaw surface. Cells with similar receptive fields were often grouped together, but adjacent skin areas were not necessarily represented in adjacent cortical patches. Thus, the cutaneous projections to this lobule terminated in a patchy or mosaic fashion.Supported by NIH grant S-R01-NS-02289     

Discharge patterns of Purkinje cells activated through the climbing fiber system by stimulation of somatic and visceral afferents

Extracellular recordings were obtained from single Purkinje cells (PC) in the intermediate part of the lobule V and VI of the cerebellar cortex in cats anaesthetized with Nembutal. A number of PC responded with a complex spike (CS) to stimulation of both the superficial radial nerve (SRN) and the vagus nerve (VN). By suppressing the transmission of mossy fiber (MF) inputs through manipulation of the level of anaesthesia, attention was paid to the simple spike (SS) activity after the CS. Modality-specific differences were found in the length of the post-CS pause, the presence of a post-pause rebound, the effect of pre-CS SS firing rate on pause duration and in the frequency and regularity of post-pause SS discharge. We concluded that these differences arouse from the location of the PC within the climbing fiber (CF) sagittal strip and from the differential activation of the inhibitory interneurons. We propose that modification of SS activity following a CS represents a possible means of information transmission by the CF system.     

The spatial organisation of climbing fibre branching in the cat cerebellum

Summary The spatial distribution of branches from climbing fibre afferents to the cerebellar cortex has been studied using electrophysiological techniques. Recordings were made from a large number of individual Purkinje cells in which climbing fibre responses were evoked by stimulation of the cerebellar surface. For each cell one or more distinct low threshold points were found from which such responses were evoked via an axon reflex pathway.For each climbing fibre studied in this way, from one to three terminations were located. These were frequently very widely distributed in the rostro-caudal direction, but were generally found to lie within a sector of the cerebellar cortical sheet which was very narrow in the medio-lateral direction.A method for estimating the conduction times in the parent axon and in each of its branches is described and applied to the experimental data relating to one axon.Quantitative measurements were made to define the parameters necessary to activate climbing fibre terminals by stimulation of the cortical surface. Using a cerebellar stimulus whose parameters were chosen on the basis of these measurements, the pattern of climbing fibre axon reflex connexions was determined over most of the accessible cerebellar surface by mapping the climbing fibre evoked potentials.Connexions formed by branching climbing fibres were found throughout the accessible cortex. They appeared always to be orientated at light angles to the long axes of the cerebellar folia, whatever the orientation of the folia with respect to the sagittal plane of the animal.     

Analysis of electrical potentials evoked in the cerebellar anterior lobe by stimulation of hindlimb and forelimb nerves

Summary Responses were evoked in the anterior lobe of the cerebellum by volleys in group I and II fibers of forelimb and hindlimb nerves — cutaneous, muscular, joint and fascial. These responses have been observed along microelectrode tracks that traverse the whole depth of the anterior lobe. These tracks have been identified in histological sections, and the recording sites along these tracks have been determined.It has been shown that there are many distinguishing features for the responses produced by the two types of afferent input to the cerebellum: climbing fibers and mossy fibers.The depth profiles are of particular importance in the differentiation of the CF and MF responses, and they correspond to those already determined for the exposed surface areas of the cerebellar cortex. As would be expected from the distribution of synapses by the CF fibers to the Purkinje cell dendrites, there is a maximum extracellular negativity deep in the molecular layer with sources superficial and deep thereto.In contrast, the mossy fiber input produces a powerful synaptic excitation in the granular layer, which is recorded there as a negative wave (N₂). The mossy fiber input by sequential relay also produces a negative wave (N₃) in the molecular layer. This wave is distinguished from the CF-evoked negative wave because it is not reversed in the fissura and the adjacent superficial molecular layer.An important distinguishing feature of the MF- and CF-evoked responses is that the latencies of the former are shorter by 6–12 msec for forelimb nerves and by 9–15 msec for hindlimb nerves. It is thus possible to measure the sizes of the MF and CF responses in the same traces.Another distinguishing feature is the failure of the CF responses with stimulus frequencies of 5–15/sec, whereas the MF-evoked potentials are well maintained above 15/sec. Also CF-evoked responses show much more size and latency variance than the MF-evoked responses, and often the facilitation of two or three volleys is required in order to evoke a stable CF response.By utilizing these various tests it is always possible to distinguish between the CF- and the MF-evoked responses recorded along the microelectrode tracks in the anterior lobe.The authors wish to express their grateful thanks to Mr. Lionel M. Davies who was responsible for the re-installation of Canberra equipment in Chicago and for its operational servicing.     

On the identification of the afferent axon terminals in the nucleus lateralis of the cerebellum an electron microscope study

Summary Axon terminals in the neuropil of the lateral nucleus can be divided into six classes, each with a specific constellation of characteristics that consistently occur together. Two of these classes have synaptic varicosities with elliptical synaptic vesicles, one in a dense, the other in a sparse matrix, and both make axosomatic and axodendritic synapses. The remaining four classes all have round synaptic vesicles and do not make axosomatic synapses. In the first of these four, the vesicles are tightly packed in a dense matrix, in another they are loosely dispersed, and in the third they are clustered. In the fourth, large granular vesicles predominate. Of these six classes, the most numerous belong to the axons of the Purkinje cell terminal arborization. These boutons resemble their counterparts in the cerebellar cortex, the recurrent collaterals of the Purkinje axon. They have elliptical and flat synaptic vesicles in a dark matrix. The varicosities terminate on somata and dendrites of large and small neurons and constitute the majority of their input. Purkinje axons constitute 86% of the total population of terminals on large neuronal perikarya and 50% of those on their dendrites, but only 78% on the somata of small neurons and 31% on their dendrites. The terminals of climbing fiber collaterals are recognized by their resemblance in electron micrographs to the terminals of the climbing fiber arborization in the cerebellar cortex. They bear round synaptic vesicles packed into a dense axoplasmic matrix and make Gray's type 1 axodendritic synapses with large and small neurons. These axons are restricted to the lateral and ventral aspects of the nucleus and constitute 5% of the terminals on large cell dendrites and 6% of those on small neurons. The axons tentatively identified as collaterals of mossy fibers are myelinated fibers with a light axoplasm containing round synaptic vesicles, dispersed throughout their varicosities. They make Gray's type 1 synapses and constitute a fair percentage of the total axodendritic contacts in the neuropil, 22% on large neurons and 28% on small neurons. The bases for these tentative identifications are discussed in detail, as are the various synaptic relationships undertaken by each class of axon. The remaining 4 classes of axons of the neuropil will be described in subsequent papers.Supported in part by U.S. Public Health Service grants NS 10536 and NS 03659, Training grant NS 05591 from the National Institute of Neurological Diseases and Stroke, and a William F. Milton Fund Award from Harvard University.     

Degeneration and electron microscope analysis of the synaptic glomeruli in the lateral geniculate body

Summary Optic fibers of retinal origin terminate in the lateral geniculate body exclusively in the so called glomerular synapses. They can be recognized on the basis of their unusually large irregular mitochondria having very few cristae. In the cat the structure of the optic terminal profiles is rather dense. The majority of terminals in most glomeruli originate from axons of other source. Relatively large axon terminal profiles of unusually light structure cannot be brought to degeneration by any interference with extraneous pathways. From Golgi information it becomes obvious that they originate from local Golgi 2nd type neurons. Small rather dense axonal profiles of the glomeruli can occasionally be traced back by degeneration to the occipital cortex (parastriate), although most of the descending cortical afferents of the lateral geniculate body terminate outside the glomeruli on more proximal parts of the dendrites. — Axo-axonic synapses are very frequent. If an optic terminal is involved, it appears that by structural standards it is presynaptic to the non optic. As judged, however, from the numerous axoaxonic contacts persisting after enucleation, many of the contacts are established between non optic axon terminals. — The progress of secondary degeneration and particularly the removal from the glomeruli of degeneration fragments is unexpectedly rapid. — The possible functional significance of these findings, especially also with regards to presynaptic inhibition, is discussed.     

Afferents to the cerebellar cortex of turtles studied by means of the horseradish peroxidase technique

Summary Origins of afferents to the cerebellar cortex from the brainstem were explored in turtles by means of the horseradish peroxidase (HRP) technique. Following relatively large injections involving all cortical layers, HRP label was observed in neural perikarya of the following structures: 1) contralateral reticular formation just lateral and ventral to the hypoglossal nucleus; 2) a few cells in the central gray of the cervical spinal cord; 3) neurons scattered in the dorsolateral, ventromedial and descending vestibular nuclei, mainly ipsilaterally; 4) a few solitary cells in the mesencephalic and medulalry tegmentum; 5) the nucleus isthmi magnocellularis caudalis on the ipsilateral side; 6) a group of small cells in the isthmic tectum; 7) the ipsilateral nucleus of the optic tract; 8) a prominent group of small cells in the isthmic region just rostral to the vestibular complex ipsilaterally. Most of these cells were localized within the so called nuclei gustatorius secundarius,-lemnisci lateralis and-isthmi parvocellularis. This parvocellular isthmic complex (PIC) was the only region containing labelled cells when small injections restricted to the molecular layer were achieved. We interpret the PIC as a source of climbing fibers, possibly corresponding to the mammalian inferior olive which migrates from the alar plate to its' ventral destination during ontogenesis. Connecting axons were sometimes homogeneously stained which permitted the tracing of connecting pathways. Contorted axon branches stained by anterograde HRP transport were found concentrated in cerebellar and superior vestibular nuclei and sparsely distributed in other vestibular nuclei.     

Activation of mossy and climbing fiber pathways to the cerebellar cortex by stimulation of the fornix in the rat

Summary The fornix of the rat was electrically stimulated with bipolar concentric electrodes to determine the properties of single unit responses in Purkinje cells of the cerebellar cortex. Both climbing (CF) and mossy fiber (MF) pathways were activated by fornix stimulation. MF responses were indicated by single or double spike responses appearing at latencies of 5–10 ms. The MF spike responses, as quantified by histogram analysis, were further identified by appearance of graded responses with increasing stimulus strength and by following at frequencies up to and greater than 20/s. CF responses were identified by characteristic complex all-or-none burst responses with latencies usually between 10 and 20 ms and with following frequencies at no faster than 10/s. Experiments which involved movement of the stimulating electrode and production of lesions around it established that the activated fiber system was within the dorsal fornix and not in adjacent areas. The results indicate that hippocampal and other limbic areas can influence the cerebellar cortex by direct mossy and climbing fiber pathways, as has been demonstrated for other afferents. It is further suggested that motor patterns linked to hippocampal activity may be regulated by this system.This work was supported by N.S.F. Grant No. 77-01174 awarded to Dr. Donald J. Woodward, and an award from the Biological Humanics Foundation     

Lateral and medial sub-divisions within the olivocerebellar zones of the paravermal cortex in lobule Vb/c of the cat anterior lobe

Summary The olivocerebellar projection to the c₁,c₂ and c₃ zones in the paravermal cortex of lobule Vb/c has been investigated in the cat using a combined electrophysiological/neuroanatomical tracing technique. The zonal boundaries in the paravermal cortex were located by recording, on the cerebellar surface, climbing fibre field potentials evoked in response to percutaneous stimulation of one or more paws. A small (10–30 nl) injection of WGA-HRP was then made either into the centre or into the medial or lateral geographical half of a chosen zone and the resultant distribution of retrogradely labelled cells within the contralateral inferior olive was plotted. The c₁ and c₃ zones were each found to consist of two mediolaterally oriented sub-zones which could be distinguished by their olivocerebellar input. The medial part of the c₁ zone received climbing fibre input from the rostromedial part of the dorsal accessory olive (DAO) while the lateral part of the c₁ zone received climbing fibre input from middle/rostral regions of the medial accessory olive (MAO). Both medial and lateral sub-zones within the c₃ zone were found to receive climbing fibre input from the rostral pole of DAO but, whereas there was heavy overlap between the olivary territories projecting to the medial c₁ and medial c₃ subzones, olivary cells projecting to the lateral part of c₃ were located more rostrally within DAO. The c₂ zone was found not to be divisible into mediolaterally oriented subzones and to receive climbing fibre input from a region of MAO located rostral and somewhat lateral to the region projecting to the lateral part of the c₁ zone. The sub-zonal organisation of the olivocerebellar projection to the c₁, c₂ and c₃ zones is discussed in relation to the functional properties of the different zones.     

Senescent pathology of cerebellum: Purkinje neurons and their parallel fiber afferents

Two groups of naive male Sprague-Dawley rats, 5-7 and 24-26 months of age, were anesthetized with continuous intraperitoneal infusion of 4% chloral hydrate. Stimulation of the cerebellar vermis molecular layer permitted measurements of 12 different electrophysiological properties of parallel fiber Purkinje cell circuitry: parallel fiber conduction velocity, refractory period, threshold, and current dependent volley amplitude; slow negative wave threshold and current dependent amplitude; Purkinje cell activation threshold, latency, and current dependent spike driving; and Purkinje cell inhibitory threshold, latency, and current dependent duration of inhibition. Old subjects demonstrated deficits on all parameters except Purkinje cell inhibitory threshold. The relevance of these findings to our previous research on senescent changes in cell number, lipofuscin deposition, and spontaneous firing of Purkinje cells is discussed.     

Topographical investigations on the climbing fiber inputs from forelimb and hindlimb afferents to the cerebellar anterior lobe

Summary Volleys in group I and II fibers of muscle nerves and group II fibers of cutaneous, joint and fascial nerves have evoked CF responses in the anterior lobe of the cerebellum. In the pars intermedia there is a fairly sharp somatotopic localization of the forelimb CF responses to the Vth lobule (Larsell) and the hindlimb to the IVth and IIIrd lobules. In the vermis there is much more admixture, with the hindlimb-evoked responses tending to dominate in the lateral vermis of the Vth lobule, and the forelimb more medially. In the IVth and IIIrd lobules forelimb responses were rare and were never large. In the medial vermis up to 1–1.7 mm from the midline there were no CF-evoked responses from the limb nerves. These distributions of CF-evoked responses are remarkably different from those reported by Oscarsson, and consideration is given to the factors responsible for this discrepancy.A more detailed examination was made of the CF-evoked responses from a large variety of hindlimb and forelimb nerves. Observations were made along many tracks usually arranged in a transverse plane, and it was found that between different recording sites along the same track or along adjacent tracks, there was a great deal of variation in the relative magnitudes of the CF-evoked responses from the different nerves. These distributions have an ill-defined patchy character so that at any focus there is opportunity for the most diverse kinds of piecemeal integration.These findings on the CF-input are considered in relationship to the mossy fiber input. It is pointed out that the pathways conveying CF-input to the cerebellum have a level of discriminative input adequate for the operation of fine control.