6 R. H. MOLE 



whicli these particular changes have occurred. Until there is some experimental 

 means for distinguishing this fraction of cells separately, there will be no 

 way of knowing if conclusions based on reproductive integrity also apply to 

 the postulated malignant transformation. 



Table I. Some somatic effects in mammals of small doses of X- or y-radiation 



(from Mole, 1962) 



The vast amount of information on chromosomal changes after irradiation 

 is also of uncertain relevance. To be observed, cells have to be dead yet it is 

 only viable cells which are relevant; if the carcinogenic change is a rare event 

 the only relevant chromosomal change w^ill be a rarely occurring one and again 

 there is no way of knowing which it is. 



HUMAN RADIATION-INDUCED LEUKAEMIA 



No species has ever been, or ever can be, examined so intensively and in 

 such large numbers as the human. Not only does this provide a sound founda- 

 tion of fact but the human evidence also covers a range of effect which is 

 inaccessible to experiment and therefore uniquely significant to any academic 

 study of leukaemogenesis. Acute leukaemia and chronic myeloid (or granulo- 

 cytic) leukaemias can be caused by radiation. There is no evidence that chronic 

 lymphoid leukaemia and other kinds of haematosarcoma can be so caused 

 but this could be merely because these conditions have a longer latent period. 



The quantitative aspects of human leukaemogenesis by radiation have 

 been thoroughly considered by many authorities (Brues, 1958; Court Brown, 

 1958; Cronkite et al, 1960; Heyssel and Brill, 1961; Murray and Hempelmann, 

 1961), especially the distinction between the linear dependence of leukaemia 

 incidence on dose (Lewis, 1957) and the curvilinear which relates incidence to 



