A carcinogenic tar applied to rabbit skin renders many more epidermal cells neoplastic than ever declare themselves by forming tumors. They may be present in large numbers and persist for a considerable time after brief tarring, yet give rise to no growths unless encouraged. The stimulus of wound healing will suffice to make some of them multiply and form tumors. No evidence has been obtained, in experiments specifically directed to the point, that the cells which tar renders neoplastic respond in this way because they are possessed of peculiarities not shared by the rest of the normal epithelium. The fact that non-specific stimulation (as e.g. wound healing) may act as the deciding influence in tumor formation brings out the need for a sharp distinction between the forces which induce neoplastic change and those which determine, or prevent, its realization in terms of a tumor. The distinction is vital to the appraisal of the many carcinogenic substances worked with nowadays. The ability of tumor cells to lie latent for long periods and respond to non-carcinogenic stimulation by multiplying into growths provides an explanation of those clinical instances in which cancer appears rapidly after acute injury to tissue that had seemed normal.