ULTRAVIOLET RADIATION AND CANCER 553 



On the other hand, the threshold for sunburn is much higher in Negroes 

 than in the white races. It is, therefore, of interest to find that cancer of 

 the skin is very rare in members of the Negro race in this country (Dorn, 

 1944), although there seems to be no evidence of a general immunity to 

 cancer in that race. Moreover, it appears that when cancer of the skin 

 does occur in Negroes it has a very different distribution, being located 

 about as frequently on unexposed as on exposed parts (Schrek, 1944a, b). 

 Here is evidence to support the idea that among the white races sunlight 

 is a principal cause of cancer of the skin. Vint (1935) reports a high inci- 

 dence of squamous-cell cancers among the natives of Kenya, accounting 

 for 36 per cent of the tumors among this Negro population. He points 

 out, however, that these tumors are associated with tropical ulcers of the 

 leg. Basal-cell cancers are rare. This evidence seems to support rather 

 than conflict with the findings on Negroes in the United States, as regards 

 etiology. Roffo (1939) states that in the Argentine all the cases of skin 

 cancer that he observed occurred in foreigners or immigrant families, none 

 in Indians or Negroes. 



Occupation. It is commonly believed that cancer of the skin is more 

 frequent in outdoor workers than in indoor workers. This is reflected in 

 such terms as "seaman's skin," "peasant's skin," and "farmer's skin," 

 which are applied to allegedly precancerous changes when they occur on 

 the exposed parts. It is obviously difficult to obtain statistics on this. 

 Perhaps the most convincing are those of Peller and Stephenson (1937) 

 and Peller and Souder (1940) who point out that mortality from cancer of 

 the skin and lip is about three times as high for the United States Army 

 and Navy as for the average population for the same age group. Other 

 evidence, some conflicting, has been discussed by the writer elsewhere 

 (1940). All in all, this part of the evidence seems the least conclusive at 

 the present time. 



Distribution. There have been many statements that the incidence of 

 cutaneous cancer is greater in those regions of the earth which receive 

 most sunlight, but there was for a long time no basis for a critical analysis. 

 In 1944, however, Dorn published extensive statistics which seem to bear 

 out this idea. He examined the incidence of various types of cancer in 

 three groups of urban areas in the United States: (1) Chicago, Detroit, 

 Pittsburgh, and Philadelphia; (2) San Francisco and Almeda County, 

 California, and Denver; (3) Birmingham, Atlanta, New Orleans, Dallas, 

 and Fort Worth. He designates these groups for his purposes as northern, 

 western, and southern, respectively, but actually they form a series of 

 nonoverlapping latitude groups. Their mean latitudes weighted accord- 

 ing to populations of the respective areas are 40°, 38°, and 32°. In Fig. 

 14-10 Dorn's figures for incidence of cancer are plotted against latitude. 

 A clear-cut north-south distribution of cutaneous cancer incidence is indi- 

 cated for both sexes of the white populations, with greater incidence in the 



