IMMUNOGLOBULIN (Ig) molecules on the surface of B lymphocytes aggregate into a 'cap'1 when the cell is treated with anti-Ig1-3. A similar process occurs with concanavalin A (con A) on neutrophil polymorphonuclear leukocytes (PMN)4. Capping occurs in two stages in both of these cell types4,5: first, the ligand-receptor complexes aggregate into multiple clusters of variable size, leaving intervening bare membrane; and second, the clusters are swept into a single mass. Both stages are temperature-dependent (that is, they do not occur in the cold), but only the latter is energy-dependent (that is, clustering occurs but capping is inhibited on cells treated with metabolic inhibitors)4-6. The formation of clusters can be attributed to the ability of the ligand to cross-link receptors5; some form of membrane activity then sweeps the clusters into the cap. Capping can occur in the absence of cell locomotion4,7, but if locomotion does occur the cap is found at the trailing end of the cell4. These aspects of capping have recently been analysed in detail4,7.