Summary.-The cell population kinetics of the transmissible venereal tumour of the dog was studied at two different stages of tumour growth using the labelled mitoses technique. At the first stage the tumours were growing with a doubling time of about 4 days; at the second stage their growth rate was limited, probably by an immune reaction on the part of the host, to a doubling time greater than 20 days. Labelling of the tumour cells was found to be extremely heterogeneous throughout the tumour. Mitotic figures, however, were present in well labelled as well as in poorly labelled fields, suggesting that thymidine did not reach all regions of the tumour nodules. The data were therefore analysed assuming that the cells in well labelled areas were representative of the total cell population in the neoplasm. The timing of the cell cycle was found to be similar in the rapidly growing tumours and in those growing more slowly. It is concluded that the slowing of growth was due to a considerable increase in the rate of cell loss as a result of the immune reaction. THE transmissible venereal tumour (TVT) is a naturally occurring coitally transmitted neoplasm of the dog which affects the external genitalia of both sexes (Karlson and Mann, 1952). It is experimentally transplantable to allogeneic hosts and was in fact used in the first recorded successful experimental transplantation of

Summary An immunoperoxidase technique has been used to detect the in vivo binding of a 2-nitroimidazole hypoxia marker in histochemical sections of a variety of excised canine tumours. The binding occurred 10-12 cell diameters away from tumour blood vessels, consistent with the expected location of hypoxic cells in tissues in which oxygen concentration gradients are established by diffusion. Hypoxic fractions ranging from 4 to 13 % have been estimated on the basis of morphometric analysis of multiple tumour sections. The binding of the marker was restricted to the cytoplasm of the cells. The marker appeared in regions adjacent to necrosis but also in regions free of necrosis. As in earlier autoradiography studies, binding was occasionally observed in cells adjacent to tumour blood vessels. Generally, binding to normal tissues was not observed. However, binding to smooth muscle cells surrounding arterioles in some sections of normal tissue and tumour tissue was observed. Viable, radioresistant hypoxic cells may compromise the effectiveness of radiation treatment of tumours (Bush et al., 1978; Thomlinson & Gray, 1955) but may also be targets for selective cytotoxins (Kennedy, 1987). In either case, the detection and measurement of tumour hypoxia would be of