Migrating distal tip cells are unusual in that the nucleus is positioned at the leading rather than the trailing edge of the cell ( Figs. 1, 2). The leading edge itself is round rather than flat or ruffling. The trailing edge ensheaths some 3-8 distalmost germ cells and the trailing lip forms a constriction around the column of germ cells. Despite such novel features, distal tip cell movement may be governed by the same cortical flow mechanism recently proposed for fibroblast and amoeboid locomotion (Bray and White, Science 239, 883 ( 1988)). In particular, a gradient of cortical contraction, lowest at the leading edge, could cause preferential cytoplasmic extension at the leading edge. Like cytoplasm, an untethered nucleus could be squeezed into this relaxed region. A prediction is that actin microfilaments producing cortical tension should be more abundant near the trailing edge of the cell and perhaps circumferentially aligned. In fact, there appears to be a concentration of circumferentially aligned microtubules in this region. How might cell direction be regulated? It is widely believed that receptors in the cell membrane couple the cortical cytoskeleton to external ligand gradients that guide the cell. If these receptors are diffusible in the cell membrane and functionless until they bind a ligand, then the front of receptor-ligand complexes that forms when the cell extends over new substratum would tend along ligand isolines ( Fig. 3). If these complexes orient new microfilaments, cells will turn up-gradient and the filaments will align circumferentially during straight segments of the trajectory. The role of the germ cells in distal tip cell migrations is somewhat complicated. They are not strictly required for hermaphrodite distal tip cell movements. Indeed, the male linker cell, which probably moves by the same mechanism, does not ensheath germ cells. Even so, the hermaphrodite distal tip cells appear to migrate faster when leading a column of germ cells. The germ cells may be serving two roles: (1) filling the cell concavity with an incompressible fluid on which to squeeze, and (2) pushing the distal tip cell from behind like a battering ram. The first role would be unnecessary in cells lacking a concavity. Passive movement of the distal tip cells caused by proliferation of the germ cells could be significant in late larvae or adults. If germ cells are not required for migration, it is not obvious why the hermaphrodite distal tip cells trouble to ensheath the distalmost cells. As the male distal tip cells apparently do not ensheath germ cells, ensheathment is not needed to regulate germ cell proliferation per se. We speculate that the ensheathment helps 'set' the cylindrical geometry of the trailing cell column. Once established, the germ cells maintain this geometry despite extensive proliferation. Cells in the somatic primordium could serve this shaping function in the male. [See Figure 1]