Charting Drosophila's non-ideal neural superposition eyes.

More than half a century ago, Buchner (1971) evaluated the visual axes of the ommatidia in a Drosophila's neural superposition eyes, using the pseudopupil method. The spatial distribution of the visual axes was reported in a stereographic projection chart. Analysis of this chart shows that the ommatidial lattice is well described by a horizontal-vertical (H-V)-coordinate system, which allows assessment of the partial interommatidial angles between the vertical columns and horizontal rows of ommatidia, Δφh and Δφv, respectively. Their angular ratio ρa = Δφh/Δφv is expected in an ideal neural superposition eye to be √3, but this appears to hold only in a central band of the eye. Dorsally and ventrally ρa is much smaller, while posteriorly it is much larger. The quite different gradients of Δφh and Δφv over the eye cause distinct regional deviations from ideal neural superposition. Instead of the interommatidial angle Δφ, considered to be the central factor determining the eyes' spatial resolving power, the local values of Δφh and Δφv will presumably be the important determinants for the spatial resolving power, at least for the vertical and horizontal motion-sensitive neurons in the fly brain. Various ways to chart the visual space of a compound eye are demonstrated and discussed.