In New Zealand malignant ovine lymphoma is a low-prevalence, sporadic disease affecting sheep 4 years and older. The aetiology of the disease is unknown but a previous study showed that when cell-free extracts of ovine lymphomas were inoculated into in utero and new-born lambs they developed a persistent lymphocytosis and cell-mediated immunity to the lymphoma extracts. Furthermore, particles interpreted as virus-like were observed by electron microscopy in phytohaemagglutinin-stimulated lymphocytes from these sheep. This thesis reports the continued investigation of sporadic ovine lymphoma including its immunological characterisation and a search for evidence of conventional retroviruses. Techniques for the detection of T cell-specific antigens, surface immunoglobulin, complement receptors and Fc receptors on lymphocytes from the blood and mesenteric lymph nodes from normal sheep were established and ranges of values for T and B cells determined. The use of α naphthyl acetate esterase (ANAE) as a specific marker for sheep T cells was investigated. No correlation was found between T lymphocytes and ANAE-positive cells. Cells from 17 cases of lymphoma were characterised for the presence of T cell antigens and surface immunoglobulin, markers for T and B cells respectively. Seven cases were T cell and 6 were B cell in origin; in one case cells displayed both markers and in another there were similar numbers of both T and B cells; the cells from the remaining 2 cases had neither marker. An apparent correlation was found between the patho-anatomical classification of the disease and the immunological origin of the lymphoma. All the T cell lymphomas were of multicentric distribution, whereas 4 of the 6 B cell neoplasms were confined to the alimentary tract and its associated lymphoid tissue. There was no correlation between the cellular morphology of the lymphomas and whether they were T or B cell in origin. Electron microscopy of several ovine lymphomas and their suspension cultures revealed only a few nonbudding particles with the dimensions of retroviruses. Smaller vesicular structures were seen associated with cells from a single suspension culture and in control cultures but these were considered unlikely to be of viral origin. Attempts were made to establish continuous cell cultures derived from ovine lymphomas. These included the culturing of lymphoma cells either over normal fibroblast "feeder" monolayers in various combinations of media and sera, or in simple Suspension cultures. Alternatively, lymphoma tissue was used in plasma clot explant cultures. Limited success occurred only with the plasma clot explant cultures where cells from one lymphoma survived 2 passages. Several fibroblast cell lines derived from the lymphoma explant cultures have been developed and passaged over 20 times. To detect the presence of retroviruses in a variety of materials derived from ovine lymphomas, 2 biochemical techniques were used. The first involved the assay of culture supernatants for 3H uridine-labelled virus in density gradients, and the second the search for RNA-dependent DNA polymerase (RDDP) activity in various preparations. Four ovine lymphoma cell suspension cultures were assayed for the production of RNA-containing virus particles. Tritiated uridine was added at the time the cultures were established and the media harvested after 96 hr. The pellets obtained following differential ultracentrifugation were centrifuged through 15 to 60 percent sucrose gradients and fractions of the gradient were assayed for acid-precipitable radioactivity. Although radiolabeled material was detected at densities of 1.15 to 1.18 gm per ml in preparations of 2 lymphoma cultures, normal lymph node cultures yielded similar results. Radiolabelled material treated with sodium dodecyl sulphate and sedimented through a sucrose gradient had a sedimentation value of approximately 7S. No high molecular weight RHA consistent with that of retroviruses was found. Further experiments using normal ovine fibroblasts led to the conclusion that the radiolabelled material detected in the 15 to 60 percent sucrose gradient was probably slowly sedimenting cellular RNA. The RDDP assay was performed on ultracentrifuged preparations from media of lymphoma and normal lymph node cultures, and Rous sarcoma and bovine leukaemia viruses were used as positive controls. Incorporation of 3H thymidine triphosphate into acid-insoluble material was detected in 4 of 16 ovine lymphoma cultures but was also found in material from control lymph node cultures. Little variation in incorporation kinetics could be evinced by altering the assay conditions, and the observed activity was not associated with a particle of density 1.15 gm per ml. Furthermore, RDDP could not be detected in preparations from the homogenates of 6 lymphomas. It was concluded that the activity observed in both the lymphoma and control lymphocyte preparations was not due to RDDP. Depressed responsiveness by lymphocytes to nonspecific mitogens has been associated with infections by retroviruses in other species. However, there were no differences in responses to phytohaemagglutinin by lymphocytes from lymphoma-inoculated and control sheep. Although conventional retroviruses have not been clearly demonstrated in association with sporadic ovine lymphoma in these experiments, the failure to detect virus does not rule out the possibility of retroviral involvement at some stage of lymphomagenesis. The development of more sensitive techniques might allow the detection of low levels of virus or viral nucleic acid sequences within cells.