The ability of a novel fish oil emulsion antigen-delivery system administered orally and by immersion, to stimulate antibody responses in the dab, was measured in a further experiment. On this occasion, oral intubation of HGG (25 mg) in saline induced no detectable responses. Immersion in a bath containing HGG in lipid emulsion induced a transient ASC response in the blood only. Anal intubation of HGG (25 mg) in saline induced a slight ASC response in the gut and blood. Oral intubation of HGG (25 mg) in lipid emulsion induced ASC responses in the gut and transiently in the gill but no response (above background) in the head kidney. None of the above methods of immunisation induced serum antibody titres. Intraperitoneal injection of HGG (1 mg) in saline induced high numbers of ASC in the head, kidney, gut and gill as well as serum antibody. The ASC response in the head kidney was detected from week 5 to 10, peaking at week 5. The response in the gill was from week 3 to 10, peaking at week 6, and the response in the gut was from week 5 to 10, peaking at week 8. The results indicate that systemic stimulation induced ASC responses in both systemic and mucosal compartments. The orally protected HGG in lipid emulsion was more effective than oral HGG in saline and anal HGG in saline in inducing ASC responses in the gut and the gill without including serum antibody, suggesting that oral immunisation can induce a common mucosal response independently of the head kidney. Leucocytes were isolated from the perfused gill of rainbow trout (<I>Oncorhynchus mykiss</I>) and fractioned on a 40-70% Percoll gradient into two subpopulations, top and bottom cell fractions. On stimulation with calcium ionophore, the isolated gill cells, following nylon wool filtration, were shown to be capable of producing chemoattractants for head kidney leucocytes at a dilution of 1:8. Only the bottom cell fraction exhibited migration toward a 2% dilution of trout serum while dilutions of 0.25% and 0.5% rainbow trout serum were not chemoattractive for either head kidney or gill leucocyte populations. The highest migration index was achieved after 1.5 h and the optimal cell number for migration was 4.65x10⁷ cells/ml. Respiratory burst activity was undetectable with isolated gill cells. Mitogenic responses of isolated and fractionated gill cells to LPS and PHA suggested the presence of few B-cells and a preponderance of T-cells.