552 RADIATION BIOLOGY 



cause such cancers. We ma^^ cite, for example, the kaiigri cancers result- 

 ing from heal, and cancers caused by carcinogenic chemical agents, such 

 as the cancer in chimney sweeps and in workers in crude petroleum. On 

 the other hand, the much more malignant, although fortunately much 

 rarer, malignant melanomas cannot be attributed to the action of sun- 

 light; these tumors parallel in distribution the pigmented nevi (Pa(;k, 

 1948), and, altliough a good many melanomas occur on the face, there are 

 other sites of predilection, particularly the genitals and the feet, where 

 sunlight would be least likely to play a part. The melanomas will be 

 excluded for the present discussion; references to cancer of the skin will 

 mean basal-cell or squamous-cell cancers. 



There have been attempts to relate the distribution of the cancers of the 

 face to the incidence of sunlight on different areas, but, as pointed out in 

 an earlier paper (Blum, 1940), these are not very convincing. Different 

 writers use different methods of describing the position of the tumors, 

 and this leads to confusion. The analysis of Magnusson (1935) who 

 related the distribution of tumors to the thickness of the epidermis and the 

 consequent penetration of light in these areas is obviously in error because 

 it fails to take into account the factor of diffusion of radiation by the 

 corneum (see Chap. 13). Altogether, such attempts seem to contribute 

 very little support to the argument that cancer of the skin is caused by 

 sunlight. But the basic finding that these tumors are limited in very 

 large proportion to the face is itself sufficient evidence to indict sunlight 

 as a causal factor, particularly when supported by the other lines of evi- 

 dence which will be discussed. There have been attempts to explain the 

 distribution of cutaneous cancer on other grounds, which do not seem too 

 convincing. Some of these have been summarized earlier (Blum, 1940). 

 Recently Corson et al. (1949) have pointed out that tumors may occur in 

 persons wearing spectacles, where these focus the light rays. They 

 attribute this to heating of the skin, but the evidence is not convincing. 

 Some glasses used for spectacle lenses transmit the carcinogenic wave 

 lengths of sunlight to a considerable extent. 



Complexion. "A commonplace in dermatological lore is that skin 

 cancer occurs more frequently in blonds than in brunets." This state- 

 ment is found in an article by Taussig and Williams (1940) in which they 

 attempt to characterize skin color and to associate it with cancer of the 

 skin. There have been numerous other attempts of this kind, but on the 

 whole such a correlation is uncertain. It is, of course, very difficult to 

 know what the terms "blond" and "brunet" mean. The idea that pig- 

 mented skin is less liable to sunburn than unpigmented skin — the fallacy 

 connected with which was discussed in the last chapter — may have had 

 considerable influence on thinking along this line. Among the white 

 races it is difficult to correlate the threshold for the erythema of sunburn 

 with complexion. 



