Of 3260 blighted tubers planted as seed in 5 successive years (1954-8), twenty-one produced sterns invaded from below ground by Phytophthora infestans (0.79 and 0.52% respectively of those infected artificially and naturally). Only in 1955 was no stem found to be invaded, in every other year they started epidemics in experimental plots. On average the first stem lesions were found 62 days after planting and 38 days before blight outbreak became general in the district. There was evidence of two patterns of dispersal for P. infestans. Distant spread was probably by air-borne sporangia, but that near to the initial sources probably resulted from sporangia transported in water. The effects of initial sources of infection on the date when the disease became general could not be traced for more than a few hundred yards. Blight often reached crops remote from the experiments at the same time as it infected every plant within them; this often happened during the weather which elicited disease forecasts. Under certain conditions of soil moisture, P. infestans from artificially infected tubers planted as seed seemed able to pass through the soil to infect leaves near it and to start epidemics without invading stems from below ground. Attacks originating in this way developed similarly to those started by invaded stems, but their importance and the conditions limiting their occurrence have yet to be determined.
