Like clinical therapy, the treatment of cancer is an experiment intended to change the course of nature. A clinician designs the experiment after reviewing the outcome of treatment given in similar situations of the past; he then makes his current therapeutic choice on the basis of what produced the most successful previous results. Until modern diagnostic and therapeutic agents became available, these routine clinical experiments could not be designed scientifically or performed effectively. Without roentgenography, endoscopy, biopsy, exploratory surgery, and suitable laboratory tests for the precise diagnosis of cancer, clinicians could not design the experiments with scientific reproducibility. Before the advent of contemporary surgery, radiotherapy, and chemotherapy, clinicians could not use the experiments effectively to alter a cancer's natural course. Because these new clinical procedures have been available for so relatively short a time—generally less than 50 years—the modern treatment of cancer is a young experimental science. Like other young