There are few data in the literature relating the behavioral sensitivity of a mammalian sensory system to the number of cortical neurons supporting that sensitivity. The present experiments show that it is feasible, although additional experiments will be necessary to define the functional relationships fully. For example, the change in the relative widths of the ODCs can be related to the numbers of V1 neurons that serve the two eyes. Hendry et al.
19 calculated the average numbers of neurons beneath a 1-mm
2 patch of V1 cortex in five normal adult cynomolgus monkeys to be 126,120 ± 5.3/mm
2, twice the neuronal density of tissue in other cortical areas. Assuming that
Macaca fascicularis and
M. mulatta have the comparable neuronal densities, in their normal hypercolumns (a left and right ODC pair) of approximately 1 mm in width,
3 each eye within a hypercolumn would have synaptic dominance over slightly more than 60,000 neurons/mm
2 (mostly monocular neurons in layer 4C). Therefore, a 32% reduction in the width of an ODC would translate into a functional loss of some ∼19,000/mm
2 neurons throughout the full cortical thickness. It is interesting that whereas monkey HT-1 showed a 32% reduction in cortical ODC width (and a presumed loss of control over as many as 19,000 neurons), there was only a minor change in CS. In comparison, the more severe blur experienced by monkeys MIK and ELL induced nearly twice the reduction in ODC space (53% and 62%, and correspondingly, an estimated loss of ∼32,000 and ∼37,000 neurons/mm
2 throughout the translaminar column) with that loss associated with a significantly greater impact on the balance in CS between the two eyes. One might expect there to be a threshold effect (i.e., that some critical number of neurons would be lost before a psychophysical change in CS could be detected using our threshold methodology). Moreover, the numbers of neurons lost before a behavioral defect could be exposed may be quite large. For example, we
20 and others,
21 have described such a phenomenon for the relationship between the loss of retinal ganglion cells and the appearance of a visual field defect in glaucoma, where as many as half the ganglion cells have become dysfunctional and died before a defect appears in the clinical visual field measurement. If a similar rationale is applied to these experiments on anisometropia, the threshold for a detectable loss in CS must be a functional loss in excess of ∼20,000 V1 neurons. With a neuronal loss greater than ∼30,000 such neurons, there is a dramatic loss in CS. Obviously, additional experiments are required to determine the full functional relationship between the numbers of V1 neurons required for a criterion level of CS. However, these few data point to the clear possibility of the quantification of the functional relationship between visual sensitivity and the number of cortical neurons.