This relationship has now been analyzed in terms of number (interacting doses) and time (interval between interacting doses). On any given diet survivorship has been demonstrated to be a function of the host's experience of the avirulent pathogen; increasing experience, either in dose or time, resulted in increasing survivorship. For a given set of interacting avirulent-virulent doses the rate at which survivorship rose was greater when the "natural" diet was fed than when the "synthetic" diet was fed. This difference in survivorship-increase rates gave rise to differences in survivorship between the two diets. These dietary differences ranged from minimal to maximal with increasing time, and then, as still more time elapsed, back to minimal again. In consequence an optimum time interval was found at which the dietary effect was maximal. This time interval was chosen for assay purposes. The assay method thus arrived at was tested repeatedly for reproducibility, and statistical analysis showed it to be reliable and capable of demonstrating a mean dietary difference in survivorship of 55 per cent.

The nature of the relationship established between avirulent and virulent S. typhimurium in mice has been discussed.