Stanley J. Schein

A role for the corpus callosum in visual area V4 of the macaque

The classically defined receptive fields of V4 cells are confined almost entirely to the contralateral visual field. However, these receptive fields are often surrounded by large, silent suppressive regions, and stimulating the surrounds can cause a complete suppression of response to a simultaneously presented stimulus within the receptive field. We investigated whether the suppressive surrounds might extend across the midline into the ipsilateral visual field and, if so, whether the surrounds were dependent on the corpus callosum, which has a widespread distribution in V4. We found that the surrounds of more than half of the cells tested in the central visual field representation of V4 crossed into the ipsilateral visual field, with some extending up to at least 16 deg from the vertical meridian. Much of this suppression from the ipsilateral field was mediated by the corpus callosum, as section of the callosum dramatically reduced both the strength and extent of the surrounds. There remained, however, some residual suppression that was not further reduced by addition of an anterior commissure lesion. Because the residual ipsilateral suppression was similar in magnitude and extent to that found following section of the optic tract contralateral to the V4 recording, we concluded that it was retinal in origin. Using the same techniques employed in V4, we also mapped the ipsilateral extent of surrounds in the foveal representation of V1 in an intact monkey. Results were very similar to those in V4 following commissural or contralateral tract sections. The findings suggest that V4 is a central site for long-range interactions both within and across the two visual hemifields. Taken with previous work, the results are consistent with the notion that the large suppressive surrounds of V4 neurons contribute to the neural mechanisms of color constancy and figure-ground separation.

A four-surface schematic eye of macaque monkey obtained by an optical method

Schematic eyes for four Macaca fascicularis monkeys were constructed from measurements of the positions and curvatures of the anterior and posterior surfaces of the cornea and lens. All of these measurements were obtained from Scheimpflug photography through the use of a ray-tracing analysis. Some of these measurements were also checked (and confirmed) by keratometry and ultrasound. Gaussian lens equations were applied to the measured dimensions of each individual eye in order to construct schematic eyes. The mean total power predicted by the schematic eyes agreed closely with independent measurements based on retinoscopy and ultrasound results, 74.2 +/- 1.3 (SEM) vs 74.7 +/- 0.3 (SEM) diopters. The predicted magnification of 202 microns/deg in one eye was confirmed by direct measurement of 205 microns/deg for a foveal laser lesion. The mean foveal retinal magnification calculated for our eight schematic eyes was 211 +/- (SEM) microns/deg, slightly less than the value obtained by application of the method of Rolls and Cowey [Experimental Brain Research, 10, 298-310 (1970)] to our eight eyes but just 4% more than the value obtained by application of the method of Perry and Cowey [Vision Research, 12, 1795-1810 (1985)].

Ganglion cell circuits in primate fovea

In the standard view all information from the fovea is relayed via only two types of ganglion cell, P (midget) and M (parasol) thought to form respectively 90–95% and 5–10% of the ganglion cell population. We characterized all 157 ganglion cells in a small patch of macaque fovea using electron micrographs of serial sections. One hundred fifteen (73%) were midget ganglion cells and were of two types, one with 28 ± 4 bipolar synapses and the other with 47 ± 3 synapses. Forty-two (27%) were non-midget ganglion cells. Most had dendrites restricted to either sublamina a or sublamina b of the inner plexiform layer, but one quarter had dendrites in both. These cells were of two types, one with input in sublamina b from blue cone bipolar cells and the other with only diffuse bipolar cell input. The ganglion cells with dendrites in either sublamina a or sublamina b were of at least one type with the possibility of more. We conclude that non-midget ganglion cells are numerous and provide additional parallel arrays to brain.