790 LIGHT AND LIFE 



action potential and found the peak of spectral sensitivity of "Rezep- 

 tors £iir das 'Dammerungssehen' " at 480 mfx. Because Antrum and 

 Schneider measured totally different responses, it is unfortunate that 

 Schneider did not repeat his experiment at higher intensities. His 

 interpretation is perhaps beclouded by the fact that Walther and 

 Dodt (58, 59) do not report that the maximum in the blue-green 

 shifts about as a function of intensity. 



The extension of the Duplicity Theory, with its numerous and 

 rather specific connotations, to the compound eyes of insects seems 

 unwarranted on the basis of present evidence. The visual systems 

 of insects present problems in abundance, and these will not be readily 

 solved by forcing meager data to fit descriptions of the vertebrate eye. 



The Morphology of Compound Eyes 



The anatomical identities of the specific color receptors of insects 

 are not known. The compound eyes of insects are composed of 

 several hundred to several thousand ommatidia. Each ommatidium 

 is a long, slender roughly cylindrical structure composed of two 

 concentric layers of cells. On the axis of the ommatidium is the 

 rhabdom, presumably the site of the primary photochemical event. 

 Surrounding the rhabdom are the cell bodies of the retinula cells, 

 approximately eight in number. The electron microscope reveals 

 that the rhabdom is composed of rhabdomeres, specialized medial 

 borders of the retinula cells. The pigment cells surround the retinula 

 cells and optically isolate the ommatidia. Distally each ommatidium 

 is capped by a corneal facet and a transparent crystalline cone. 

 Proximally each retinula cell is believed to send an axon through 

 the basement membrane into complex synaptic relations wdth nerve 

 cells of the optic lobe. 



In 1826 Johannes Midler fornudated the mosaic theory of vision 

 by compound eyes in which he suggested that the individual om- 

 matidia respond as units. This theory has had general acceptance 

 since the work of Exner (12) . Recently, however, de Vries and Kuiper 

 (53) report that the visual angle of an ommatidium is about eight 

 degrees, several times wider than previously supposed. These authors 

 also present optical evidence that the individual rhabdomeres of an 

 ommatidium (of a fly?) can be separately illuminated, and they sug- 

 gest that the basic receptor unit may be the retinula cell rather 

 than the ommatidium. On morphological grounds this argument is 

 reasonable for flies. The rhabdomeres of dipterans are anatomically 



