BERGE: Isee. Isee. Okay. Well, while you were talking, I took some notes and I was wondering if I could go back and ask you aboutthose. KOHN: Sure. BERGE: When you weretalking about your time at the Shields Warren Radiation Lab, you mentionedthestatistical analysis of radiation therapy treatment. Were the treatments in the various hospitals experimental, or were they by that time already established therapies? KOHN: They were not experiments in the conventional sense. BERGE: So what wasthe necessity at that time of doing thestatistical analysis. KOHN: Cure rates vary with (1) the type of cancer and (2) the stage of the disease at which treatmentis initiated. To make comparisons between different cancers, or different stages of cancer, it is necessary to makestatistical comparisons. Likewise when seeking to improve the cure rate by some changein the treatment schedule, for example, by increasing the dose by ten percent or by giving five treatments a week instead of three. BERGE: Howdid they establish a dose, originally, for example, like a five times a week? Let's say it was five times a week. KOHN: In the early days of radiation therapy, it was generally held, especially by the Germanschool, that the biggest dose (tolerated) given as fast 23