ULTRAVIOLET RADIATION AND CANCER 531 



or by any of the colored ones, did not. In addition, Rolfo carried out 

 similar experiments with sunlight; he is, in fact, the only experimenter to 

 have induced cancers with natural sunlight. h\ this case, too, "verre 

 transparent" filters prevented the formation of tumors. If the assump- 

 tion is correct that Roffo's "verre transparent" was ordinary window 

 glass, he may be credited with having discovered that the carcinogenic 

 wave lengths are those shorter than about 0.32 /x,^ i.e., that the carcino- 

 genic ultraA'iolet has the same long-wa\'e-length limit as the erythemal 

 spectrum. The question was definitely settled a little later. Funding 

 et al. of Copenhagen in 1936 reported extensive experiments on the car- 

 cinogenic wave lengths before the Comite International de la Lumiere at 

 Wiesbaden. They found that a filter which transmitted wave lengths 

 0.28 fx and longer if placed in front of a mercury arc permitted induction of 

 cancer, but that window glass cutting off the 0.313 n and all lines of 

 shorter wave length prevented induction of the tumors.^ In 1941, Rusch 

 et al. of Wisconsin reported similar experiments with like results. They 

 were unaware of the earlier work of Funding et al., which was not readily 

 available. The experimental evidence serving to establish the long-wav^e- 

 length limit of the carcinogenic radiation is summarized in Table 14-1, 

 which includes a few additional observations. 



Up to the present, no studies have been reported in which monochro- 

 matic radiation was used to induce cancer, other than some by Rusch 

 et al. (1941) that seem inconclusive and some by Blum and Lippincott 

 (1942) who used the 0.2537-m line from a low-pressure mercury arc and 

 showed that this wave length is weakly effective in inducing cancer in 

 albino mice (see Table 14-1). The idea seems current that studies with 

 monochromatic radiation should show much more than is already known 

 about the carcinogenic process, and that studies with polychromatic radi- 

 ation are faulty in some way not clearly specified. The difficulties of 

 working with monochromatic radiation should become obvious when it is 

 realized that groups of identically treated mice as large as 40 in number 

 are needed if reasonably good comparisons are to be made. The difficulty 

 of exposing such numbers of animals to monochromatic radiation must be 

 obvious. But, besides the experimental infeasibility, there are appar- 

 ently insurmountable difficulties of interpretation that would arise once 

 the studies with monochromatic radiation were completed. An analysis 

 by the writer (1943a), based on a few experiments with filters, indicates 

 the uncertainty and the apparent futility of more exact measurements 

 with monochromatic radiation. The situation may be compared with 

 that of sunburn, discussed in Chap. 13. The erythemal spectrum has 

 been determined with monochromatic radiation, and a smooth curve may 

 be drawn to describe it, whereas we have only a rough idea of the shape of 



' Approximately the shortest wave lengths transmitted by ordinary window glass. 

 *See Fig. 13-3, Chap. 13, for the position of the mercury lines. 



