Although the aetiological association between soot and cancer of the scrotum was recognized by Percival Pott as early as 1775, the role of mineral oils in the production of skin cancer was not recognized until considerably later. Joseph Bell of the Edinburgh Royal Infirmary reported a case of carcinoma of the scrotum in a Scottish shale oil worker in 1876. Southam and Wilson (1922) pointed out the high incidence of scrotal cancer in Lancashire cotton mule spinners and, although various theories of causation were advanced, the production of tumours in mice by the cutaneous application of lubricating oil by Leitch (1924) indicated that some constituent of the spindle oil was the carcinogen. The carcinogenicity of a large number of petroleum oils was investigated by Twort and Twort (1931) who showed that many were carcinogenic. The high incidence of skin cancer in Scottish shale oil workers was described by Scott (1922). Experimental tumours were produced in mice by applying shale oil to the skin (Leitch, 1922), and finally Berenblum and Schoental (1943), by using the Chromatographie method of separation, isolated 3-4 benzpyrene from this oil and showed that there were other carcinogens present. The carcinogenic agents in petroleum oils, however, probably differ qualitatively and quantitatively from those in shale oils which are produced by destructive distillation. So far no pure carcinogens have been isolated from petroleum oils. In 1920, it was made obligatory to notify occupa tional skin cancer due to paraffin or mineral oil or any of their compounds, residues or products, contracted in a factory or workshop. These notifi cations have been invaluable for finding out the extent to which the disease occurs, and in con junction with the examination of death certificates, have furnished the basis of the extensive researches of S. A. Henry (1946; 1947). This work has empha sized the high incidence of skin cancer in cotton operatives, but in the engineering industry relatively few cases have been observed. For example, in the period 1911-38, only 71 metal workers (including precious metals) died with scrotal cancer out of a total of 715 such deaths recorded among workers exposed to mineral oil, the bulk of the cases having occurred in the cotton industry. The incidence in the engineering industry appears even lower when notifications instead of deaths are considered, but, as is generally recognized, notifications are unreliable until a definite hazard is widely appreciated. Our studies indicate that the hazard of skin cancer is to