It is now generally agreed that a cancer represents the ultimate phase of a definite series of changes rather than a mere fortuitous incident in the life of the tissues, and, although the nature of these changes is not yet completely understood, it is known that they result from developmental change, from reaction to prolonged irritation or from unsuccessful attempts at tissue repair. Evidence giving support to this has been accumulated by investigators in three separate fields—experimental workers, pathologists and clinicians. EXPERIMENTAL WORK Experiments have been carried out chiefly on the so-called tar cancer in mice. Cramer,1 by applying one of the known carcinogenic substances to the skin of mice, was able to produce typical epithelioma in these animals. He observed that the process is not continuous but is composed of two phases: (a) a process of long duration inducing in the cells a condition of "potential" malignancy which