Item

Abstract

The compound eye of the housefly Musca domestica L. contains two different types of receptors. The visual acuity of the eye is determined by the divergence angle Δϕ between the optical axes of neighbouring ommatidia. Δϕ and its dependence on the mean pattern brightness is determined by an evaluation of the optomotor responses elicited from various test patterns. Based on the assumption that the visual fields of both types of receptors approximate the shape of a spatial Gaussian distribution they can be characterized by their half-width, designated as the acceptance angle ΔQ. The contrast transfer from the optical environment onto the receptor cells is limited by ΔQ. It is shown experimentally that ΔQ depends on the mean environmental brightness. The characteristic values Δϕ and ΔQ constitute the limiting factors for the light flux received by the receptors. The light flux Φ exciting the receptor cells is proportional to (ΔQ·Δϕ)2. If the product ΔQ·Δϕ is kept constant, there exists a certain ratio ΔρΔφ that leads to an optimal combination of both, contrast transfer and resolution. The ratio ΔρΔφ

is experimentally determined and compared with the optimal condition. The torque exerted by fixed flying Muscae has been used as a measure of the reaction strength of the optomotor response elicited by the rotation of cylindrical patterns consisting of periodic distributions of surface brightness. The responses were investigated under different spatial wavelengths, contrasts, contrast frequencies and mean pattern brightness. Detailed results are:

1.

The visual acuity (optical resolution power) of the compound eye of Musca is determined by the divergence angle Δ ϕ between the optical axes of those adjacent ommatidia which are not positioned in the same horizontally oriented row but — closer together — in neighboured rows.
2.

Δϕ and consequently also the visual acuity do not depend on the mean environmental brightness.
3.

The acceptance angle ΔQ changes with the mean brightness of the environment. According to experimental conditions only the minimal acceptance angle Δ min can be experimentally determined. Δ min decreases with increasing mean pattern brightness from 3.6°–4.1° to 1.7°.
4.

The decrease of ΔQ min with increasing mean pattern brightness is not caused by a change of the acceptance angles of single receptors. The present tentative explanation is that the centrally located receptors No. 7 and 8 are participating in the uptake of relevant visual information at a critical brightness level.
5.

Near the optomotor threshold the large acceptance angle ΔQ min=3.6° at very dim light would thus be associated with the receptors No. 1 to 6, whereas the smaller acceptance angle ΔQ min=1.7° with the receptors No. 7 and 8.
6.

Due to a sample spacing of Δϕ=2°, the acceptance angles of neighbouring receptors No. 1 to 6 show a considerable overlap.
7.

Based on anatomical data, the difference in absolute light sensitivity for both receptor systems is calculated. It is predicted that the absorption rate of light quanta in the less sensitive system of the receptors No. 7 and 8 should be reduced by a factor of 24–48 compared to the more sensitive system of the receptors No. 1 to 6. This factor nicely meets the experimentally determined brightness thresholds of both receptor systems.
8.

The optimal condition ΔρΔφ

is nearly fulfilled by the receptor system No. 7 and 8 of Musca. The experimentally determined ratio amounts to ΔρΔφ=0.83. For the receptor system No. 1 to 6 one finds ΔρΔφ

=1.86; in that system the transfer of spatial wavelengths is mainly limited by the reduced contrast transfer which drops to low values before the optical resolution limit is reached.
9.

Based on the hypothesis that movement perception of the fly Musca is due to a correlation of sensory data one would expect an optomotor peak reaction at a spatial wavelength of λ max=8° and a decrease of the optomotor response towards longer spatial wavelengths. The experimental data are in conflict with these predictions. The present notion is that the absence of the expected reaction decrease is not likely to be caused by a saturation effect in the reaction but rather is explainable in terms of a receptor system consisting of larger numbers of receptor types No. 1 to 6 whose excitations being summed before a correlation evaluation takes place.