Choroidal melanomas may remain localized for more than 30 years (Newton, i965); malignant melanomas of the iris may not metastasize at all (Ashton, i964), and some of these intraocular neoplasms may undergo spontaneous regression (Fuchs, I91o). From time to time several hypotheses have been advanced in an attempt to explain these puzzling features of the behaviour of uveal melanomas. Furth (I953) first proposed that melanomas are tumours conditioned by changes in host environment. Dunphy (I 957) then elaborated the theory of "defensive substances" present in the body which prevented a uveal melanoma from disseminating into the general circulation. Anderson and O'Neill (I 957) referred to these substances as "biochemical barriers". This concept was supported by Reese (i963), who stated that certain "unknown factors", probably immunological in nature, kept the tumour dormant for a long time or inhibited their growth completely, metastases developing only when there was a breakdown in the immunity of the host. Jensen (I 963), in his thesis on uveal melanomas, upheld the views of earlier workers and argued that variations in metastases, such as their frequency, temporal relationship, latency, and in certain cases spontaneous necrosis, suggested the existence of such factors as immunization and resistance. Duke-Elder and Perkins (i 966) appeared to favour this hypothesis, thus explaining infrequent and delayed metastases in spite of the early dissemination of tumour cells in the blood stream. Such a latency is not peculiar to ocular melanomas and in fact several instances of dormant cancers of the breast, kidney, etc., are on record (Gordon-Taylor, I 959). To explain dormancy and spontaneous regression, Burnet (i 964, I967) suggested that the homograft rejection phenomenon was essentially a homeostatic mechanism to eliminate cells of any clone undergoing malignancy, and since a neoplastic cell carried a new surface antigen it might be destroyed by a similar mechanism to graft rejection. It is now believed that cancers probably develop and proliferate as a result of a failure in the equilibrium between the spontaneous proliferation of neoplastic cells and the ability of the body's immune mechanism to control this process (Turk, I 969). The role of immunological processes in the control and development of malignant neoplasms has been appreciated only in the last few years, although it has been suspected since the work of Simon and Thomas (I908). Foley (I953) is considered to be the first to have demonstrated specific antigenicity of chemically induced tumours. This was most closely anticipated by the work of Gross (I943) who demonstrated active immunization in C3H mice to transplants of early methyl-cholanthrene-induced tumours.
