The lateral eyes of Crustacea and Insecta consist of many single optical units, the ommatidia, that are composed of a small, strictly determined and evolutionarily conserved set of cells. In contrast, the eyes of Myriapoda (millipedes and centipedes) are fields of optical units, the lateral ocelli, each of which is composed of up to several hundreds of cells. For many years these striking differences between the lateral eyes of Crustacea/Insecta versus Myriapoda have puzzled evolutionary biologists, as the Myriapoda are traditionally considered to be closely related to the Insecta. The prevailing hypothesis to explain this paradox has been that the myriapod fields of lateral ocelli derive from insect compound eyes by disintegration of the latter into single ommatidia and subsequent fusion of several ommatidia to form multicellular ocelli. To provide a fresh view on this problem, we counted and mapped the arrangement of ocelli during postembryonic development of a diplopod. Furthermore, the arrangement of proliferating cells in the eyes of another diplopod and two chilopods was monitored by labelling with the mitosis marker bromodeoxyuridine. Our results confirm that during eye growth in Myriapoda new elements are added to the side of the eye field, which extend the rows of earlier-generated optical units. This pattern closely resembles that in horseshoe crabs (Chelicerata) and Trilobita. We conclude that the trilobite, xiphosuran, diplopod and chilopod mechanism of eye growth represents the ancestral euarthropod mode of visual-system formation, which raises the possibility that the eyes of Diplopoda and Chilopoda may not be secondarily reconstructed insect eyes.
