The optic lobes of Lepidoptera.

Variants of the Golgi-Colonnier (1964) selective silver procedure have been used to show up neurons in insect brains. Neural elements are particularly clearly impregnated in the optic lobes. Three classes of nerve cells can be distinguished; perpendicular (class I), tangential (class II) and amacrine cells (class III). There are many types of neurons in each class which together have a very wide variety of form. Their components are related to specific strata in the optic lobe regions. Short visual cells from the retina terminate in the lamina in discrete groups of endings (optic cartridges). Pairs of long visual fibres from ommatidia pass through the lamina and end in the medulla. Class I cells link these two regions in parallel with the long visual fibres and groups of these elements define columns in the medulla. These in turn give rise to small-field fibres that project to the lobula complex. Tangential processes intersect the parallel arrays of class I cells at characteristic levels. Some are complex in form and may invade up to three regions. Another type provides a direct link between the ipsi- and contralateral optic lobe. Amacrine cells are intrinsic to single lobe regions and have processes situated at the same levels as those of classes I and II cells. A fifth optic lobe region, the optic tubercle, is connected to the medulla and lobula and also receives a set of processes from the mid-brain. There are at least six separate types of small-field relays which could represent the retina mosaic arrangement in the lobula.