268 L. H. Gray 



processes can be destroyed by a sufficiently large dose to either 

 nucleus or cytoplasm. The mammalian erythrocyte provides 

 positive evidence for cytoplasmic damage (Sheppard and 

 Stewart, 1952). In ten of the cases listed in Table I it has been 

 found, however, that over a certain range of doses a given 

 effect is produced if the nucleus is included in the field of 

 irradiation and not produced if the nucleus is either excluded 

 from the field of irradiation or introduced into the cell after 

 irradiation of the cytoplasm. In one or two cases the same 

 effect has been achieved by the irradiation of cytoplasm 

 without irradiation of nuclear material, but only by the use 

 of doses 3-20 times greater than those which are sufficient 

 when nuclear material is irradiated. If we admit also the less 

 direct evidence presented in the case of lethality in haploid 

 and diploid yeast and in Neurospora conidia, then in thirteen 

 cases out of sixteen it may be said that the cell owes its 

 sensitivity predominantly to the susceptibility of its nuclear 

 material to injury by ionizing radiation. 



Account must be taken of the fact that in all the cases con- 

 sidered, the nucleus has been irradiated in the presence of 

 cytoplasm. It might be argued, therefore, that the injury 

 which makes itself apparent in the nucleus is in fact secondary 

 to a cytoplasmic injury. 



It would appear that nuclear damage can be subdivided into : 



(a) Forms in which the damage is seen in most or all cells 

 at a given stage of development, is graded with dose, 

 and for which formal dose relations are not character- 

 istic of a single particle initiation. Cleavage delay, 

 division delay and unspecific nuclear pyknosis are 

 typical examples. 



(b) Effects seen in some cells and not at all in others, such 

 as the induction of mutations and chromosome struc- 

 tural damage, having dose relations characteristic of 

 individual particle effects. 



With regard to Class (b) it would, in my view, be difficult to 

 sustain this hypothesis in the light of the experiments of 



