ONE of the best ways to investigate the specific properties of the cell surface is by immunohistochemistry. For this antibodies are required which react selectively with the cell surface antigens. Rabbit serum against normal mouse liver cell “ghosts” (NLCG) has been shown to stain specifically the membranes of parenchyma cells1. It was not clear whether this staining was caused (at least partly) by the presence of organospecific antigen(s) on the cell surface. If this is so, then we wonder whether this antigen(s) is present at the surface of hepatoma cells.