MANNER OF PRODUCTION OF MUTATIONS 511 



which growth had been allowed to go on actively, by reason of their 

 growing tips having been shielded from the radiation, and none in the 

 other 400 colonies parasitizing plants in which growth had been inhibited 

 by inclusion of the growing tips in the exposure. They believe this 

 grouping to indicate that the mutations occurred by misconstruction of 

 daughter virus particles, caused by incorporation within them of altered 

 substrate that had been synthesized only when growth of the host plant 

 was allowed. Although these results, which have been repeatedly cited 

 in evidence of transverse induction by ionizing radiation, do appear sug- 

 gestively similar to those later obtained by Stone, Wyss, and others with 

 ultraviolet (see pp. 548-552), more extensive data along these lines are 

 needed before the effect in question (and still more before the hypotheses 

 here proposed) can be regarded as well founded. 



7. EVIDENCE OF TIME LIMITATION IN MUTAGENESIS 

 BY IONIZING RADIATION 



In order that evidence might be obtained on the question whether the 

 mutagenic action of radiation continues long after the exposure has 

 ceased, as it might do if long-lasting mutagens were produced or if some 

 of the genetic material itself were brought into a less stable state (as in 

 Erwin Baur's spontaneous cases of "premutation"), the first experiments 

 showing the production of mutations by ionizing radiation (Muller, 

 1928b) had included tests not only of the mutation frequency existing in 

 the exposed individuals but also of that in their descendants of the first 

 and second generation after treatment. 



Some slight heightening of the frequency of origination of lethals was 

 indeed found to occur in the first generation after treatment, in those 

 chromosomes which were descended from the treated spermatozoa. 

 However, this result was interpreted (correctly, as later work indicated) 

 as being due to some of the lethals having arisen as " fractionals " (mosa- 

 ics) in only one of two sister chromatids immediately derived from the 

 irradiated sperm. Since these lethals were probably connected with 

 structural changes, this result did not necessarily imply any aftereffect, 

 except in so far as the union of broken ends is postponed until after 

 fertilization. At that stage the fragments of one chromatid derived 

 from a broken chromosome could join in a manner such as to give a 

 lethal effect, while those of the sister chromatid could join up in a different 

 manner or undergo restitution, so as to be nonlethal (see p. 377). The 

 tests were of such a nature that those fractional (approximately half and 

 haK in most cases) individuals of the first generation after treatment, 

 which had the lethal-bearing chromatid present in some of their germ 

 cells and the nonlethal chromatid in other germ cells, would not have been 

 recognized as having the lethal since some of their sons, those receiving 



