P(1-11).p are all born in L1 animals. In hermaphrodites, P(1-2).p and P(9-11).p fuse with the large hypodermal syncytium, hyp7, shortly after they are born (White, Anness and Thomson, WBG July 1980; Kenyon 1986). In the absence of an anchor cell, P(3-8).p divide once in L3 animals and then fuse with hyp7 (the 3 fate). In the presence of the anchor cell, only P(3-4).p and P8.p fuse with hyp7 (in L3 animals after dividing), with P(5-7).p giving rise to the vulva. We suggest, from our work on lin-15 mosaics (see accompanying abstract), that lin- 15, and perhaps other Muv genes, could encode components of machinery within hyp7 required for the fusion of new cells to hyp7. (The simplest version of this idea is that the lin-15(+) expression is needed only in hyp7 and not in the captured cells.) According to this view, the ground state of an unfused Pn.p cell in both hermaphrodites and males (see below) is 1 (or 2 , depending on other interactions). In wild type, the immediate fusion of the daughters of P3.p, P4.p, and P8.p prevents further cell divisions. The anchor cell signal, however, allows the P(5-7).p daughters to override the cell fusion program and to complete 1 or 2 lineage patterns. In lin-15 mutants, the daughters of P3.p, P4.p and P8.p may complete 1 or 2 lineage patterns as a direct consequence of failing to fuse with hyp7. This suggestion is consistent with the effect of lin-15 mutation in generating ventral protrusions in males. In wild-type males, P(7-8).p fuse soon after they are generated, but P(3-6).p do not fuse until much later (Kenyon 1986). We have observed that only P( 3-6).p appear capable of extra divisions in lin-15 males, as if it is the fusion of these latter cells that is defective. It is possible that lin-15, and related genes, may affect other hypodermal fusions occurring during embryogenesis and larval development. Indeed, a second effect of lin-15 mutation is to transform P11.p into P12.p in hermaphrodites (Fixsen et al. 1985). In wild type, P11.p fuses with hyp7 in the L1 while P12.p divides once more to generate a mononucleate hypodermal cell and a sister cell programmed to die . Perhaps fusion of P11.p to hyp7 fails when lin-15( +) activity is low, permitting the cell to express the P12.p program. Interestingly, Ferguson and Horvitz (1985; 1989) have shown that all extant lin-15 alleles are probably hypomorphs and that the gene may be essential. (They have also shown that under certain conditions one allele exhibits a maternal effect for the Muv phenotype.) We suggest that it may be worthwhile to look for hypodermal fusion defects in various Muv and Vul mutants. For example, some Vul mutations could act by causing premature fusion of P(3-8).p.