Tony Hyman

Q & A the major classes and many physical features are still under debate. To polarise arthropod trees, and define the likely morphological characters and developmental mechanisms of the arthropod ancestor, analysis of sister phyla in the Panarthropoda is required. Tardigrada are all very small, likely to have highly modified developmental processes, and have highly idiosyncratic morphology, and so may be of limited use in this regar Onychophora have provided fertile ground for investigation. Examination of velvet worm body patterning has shown that the animals have a mix of segmental and non-segmental features. Externally, the obvious segmental arrangement of the limbs is matched by some other structures (such as excretory organs) but not by others (for example, there are no obvious segmental borders in the body wall). Some segmentation genes first identified in insects are implicated in velvet worm segmentation (as they are in annelids), but overall their development is strikingly different from that of arthropods. For example the arthropod (and tardigrade) post-cephalic nervous system is characterised by paired ventral nerve cords linked by segmental ganglia; in Onychophora, evidence for segmenta ganglia is absent — there are paired ventral cords, but no segmental patterning other than that imposed b the presence of the serially repeated legs. The tripartite arthropod brain was thought to have a counterpart in Onychophora, but recent cell-level and developmental analyses have shown that, while the protoand deuterocerebral regions have neural input from antennae and jaws, no putative third part innervating the slime papillae could be identified. The tritocerebrum thus seems to be an arthropod innovation.