MANNER OF PRODUCTION OF MUTATIONS 491 



breakage, there were more cases in which the broken ends had failed to 

 unite with any others, either by isochromosome formation or by union 

 with ends derived from other, nonhomologous breaks. 



The lesser frequency of union of broken ends caused by a-ray treat- 

 ment has been taken (Lea, 1946; Catcheside, 1948) to mean that the 

 broken ends had, in the case of such radiation, more often been so affected 

 as to make them actually incapable of union, i.e., that they had been 

 caused to "heal." However, just as radiation giving relatively long dense 

 clusters of ions is, as has just been noted, especially likely to cause two 

 breaks in adjoining chromatids, so it must be especially likely also to 

 cause two or more nearby breaks within the same chromatid. The 

 more of these nearby breaks that occur in a given chromatid the smaller 

 will be the chance that one of the major chromosome pieces thus formed 

 will manage to undergo union with another major piece, either directly 

 or via an interstitial minute fragment. At the same time it is to be 

 expected that these interstitial fragments would often be too small to be 

 readily detected by ordinary cytological means. Since the a-ray tracks 

 are characterized by especially dense ionization, even as compared with 

 that in the proton tracks produced by neutrons and in the "tails" of 

 electron tracks, and since a tracks are also far longer than these tails, 

 a radiation would produce an unusually high frequency of neighboring 

 breaks in a given chromatid or chromosome. It is therefore unnecessary 

 to suppose that this radiation is unusually potent in causing the ends to 

 become actually unjoinable (healed). 



The question of whether one ionization or activation is sufficient to 

 cause a break or whether a cluster is necessary might be decided if suit- 

 able studies could be made of the comparative effectiveness, in producing 

 individual breaks, of radiation of different specific ionizations applied at 

 a stage when the threads are known to be really single. Data of Thoday 

 (1942), according to calculations of Lea and Catcheside (1942), show that 

 neutrons are about twice as effective as ordinary X rays in causing chro- 

 mosome breaks that can later be detected, when applied to Tradescantia 

 microspores at a stage which does not give detectable breaks in individual 

 chromatids. This result would appear to show that ionizations or activa- 

 tions which are close together act synergistically in causing breaks of 

 single threads. However, it is still possible that the chromosomes con- 

 sisted of two or several closely adjacent sister strands at the time they 

 were treated, even though in breakage these strands behaved as though 

 they were single, in that the sister strands could not break, or at least 

 remain broken, separately, and that they also joined sequentially and 

 never by a fusion of a broken end with its side-by-side sister end. In 

 that case, as when visibly separate chromatids were treated, a break 

 would be expected to be more likely (or only able) to persist if it involved 



