SUMMARY1. Cat retinal ganklion cells may be subdivided into sustained and transient response-types by the application of a battery of simple tests based on responses to standing contrast, fine grating patterns, size and speed of contrasting targets, and on the presence or absence of the periphery effect. The classification is equivalent to the 'X'/' Y' (linear/nonlinear) subdivision of Enroth-Cugell & Robson which is thus confirmed and extended.2. The sustained/transient classification applied to both on-centre and off-centre cells.3. Lateral geniculate neurones may be similarly classified by the same tests. Occasional concentrically organized cells had a mixture of sustained and transient properties.4. A technique for simultaneous recording from a geniculate neurone and one or more retinal ganglion cells providing its excitatory input showed that the connexions were specific with respect to the sustained/ transient classification as well as the on-centre/off-centre classification. Most geniculate neurones are excitatorily driven only by retinal ganglion cells of the same functional type. In a few cases the inputs were mixed but only with respect to the sustained/transient classification.5. Sustained retinal ganglion cells had slower-conducting axons than the transient type. The same was true for lateral geniculate neurones but in this case the distributions showed considerable overlap.6. The sustained/transient classification is the functional correlate for the well-known segregation of optic nerve fibres into two conduction groups.7. The pathways carrying sustained and transient information remain essentially separate from retina through the lateral geniculate nucleus to the striate cortex.

SUMMARY1. Nine hundred and sixty cat retinal ganglion cells were evaluated with respect to receptive-field organization and latency to antidromic activation of their axons from optic-tract and mid-brain positions.2. The vast majority (92 %) had the familiar concentric centre/surround organization. As in earlier work these could be classed as sustained or transient, independently of the centre type. About 13 % of the concentric cells were characterized by relatively sluggish responses to conventional visual stimuli which yielded brisk responses from the others. The sluggish cells constituted a previously unspecified class of concentric receptive fields.3. The responses of brisk and sluggish cells to a variety of stimuli were described with a view to developing procedures for distinguishing them on functional grounds.4. Measurements of latency to antidromic activation of retinal axons confirmed earlier work in showing that cells classed as brisk-transient had the shortest conduction times from the optic tract. Cells classed as brisksustained had intermediate conduction times and from earlier work would constitute an important input to the lateral geniculate nucleus. A proportion of the brisk-sustained axons reached the pretectal region (especially on-centre types) and a small minority reached the superior colliculus (especially off-centre types).5. Sluggish cells had generally slower antidromic conduction times; despite some overlap with the brisk-sustained class, the slower conduction provided independent support for the functional differentiation. Sluggish axons reached the pretectal region and superior colliculus. 6. The brisk-sustained cells constituted the majority of the recordings in the area centralis.

1. Two groups of interneurons that are involved in the organization of the lateral geniculate nucleus (LGN) are described. The cell bodies of one group lie within the LGN; these units are referred to as intrageniculate. The cell bodies of the other group are found immediately above the LGN at its border with the perigeniculate nucleus; these units are referred to as perigeniculate. 2. Intrageniculate interneurons have center-surround receptive fields that resemble those of relay (principal) cells. They can be subdivided into brisk or sluggish and sustained or transient categories. They are stimulated transsynaptically from the visual cortex and have a characteristic variation in the latency of their spike response to such stimulation both at threshold and for suprathreshold stimuli. The pathway for this stimulation appears to be via cortical efferents to the LGN. Intrageniculate interneurons receive direct, monosynaptic retinal inputs, as determined by recording simultaneously from such interneurons and from the ganglion cells which provide excitatory input to them. Similar to relay cells, they are shown to have one or two major ganglion cell inputs. 3. Perigeniculate interneurons are generally binocularly innervated and give on-off responses to small spot stimuli throughout their receptive field. They respond well to rapid movement of large targets. They respond to electrical stimulation of the retina with a spike latency that falls between that of brisk transient and brisk sustained relay cells. This latency is one synaptic delay longer than that of brisk transient relay cell activation and suggests that they are excited by axon collaterals of these relay cells. Electrical stimulation of the visual cortex is also consistent with this model; the latency of the response of perigeniculate interneurons is approximately one synaptic delay longer than the latency of the response of brisk transient relay cells. 4. The interneuronal pathways described are consistent with proposed circuits that subserve the generation of IPSPs that arise in response to optic nerve and visual cortical stimulation. We now show that such inhibition has feed-forward (intrageniculate) and feed-back (perigeniculate) components that are mediated by two different classes of geniculate interneurons. It is suggested that the intrageniculate interneurons are involved in precise, spatially organized inhibition and that the perigeniculate interneurons are part of a more general, diffuse inhibitory system that modulates LGN excitability.

No abstract

No abstract