Therapeutic tactics under development to restore sight to the blind rely on stimulating elements in the human retina where more than 50 different types of neuron are organized into specialized circuits, some of which mediate conscious perception, while others serve non-imaging forming functions, such as circadian photoentrainment and guiding motor movements. However, there is little agreement about which of the more than 20 ganglion cell types are responsible for specific visual functions. Similarly, little is understood about the retinal circuits contributing to the diverse tasks outside the human perceptual experience. These represent fundamental gaps in knowledge that must be filled to provide information necessary to guide efforts to preserve or restore visual functions such as object recognition without inadvertently interfering with essential non-image- forming functions. Thus, the long-term objective of the work proposed here is to fill these gaps in knowledge by making tangible progress on two key issues. First, the ability to see and recognize objects requires a diversity of cells carrying different information to specific brain areas involved in conscious vision. For example, the shapes and boundaries of objects in a visual scene are represented by the percepts of black, white, shades of gray, and color. Specific Aim 1 is to test the hypothesis that previously unappreciated features of the midget circuitry, both in the inner and outer plexiform layers, carry out the initial computations responsible for color and black and white vision, and that this is done by circuitry that separates midget ganglion cells into six types corresponding to the sensations of red, green, blue, yellow, black, and white, each of which are associated with signaling surface reflectance, not merely the spectral properties of light in the retinal image. Second, 3D reconstructions of primate retina using serial block face scanning electron microscopy (SBFSEM) have revealed an unexpected diversity and quantity of retinal ganglion cell types carrying short-wavelength (S) cone signals. The S-cone connectome includes redundant color-coding ganglion cells that project to many brain areas providing the same color information for a broad range of functions besides conscious color vision. Specific Aim 2 is to gain a complete understanding of all the types of retinal ganglion cells carrying S-cone signals in the primate retina, the circuitry responsible for their response characteristics, and the role each plays in guiding our behaviors. This will be done using SBFSEM and single-cell electrophysiology.